Calcium-activated potassium channels and endothelial dysfunction: therapeutic options?

Abstract

The three subtypes of calcium-activated potassium channels (K_Ca) of large, intermediate and small conductance (BK_Ca, IK_Ca and SK_Ca) are present in the vascular wall. In healthy arteries, BK_Ca channels are preferentially expressed in vascular smooth muscle cells, while IK_Ca and SK_Ca are preferentially located in endothelial cells. The activation of endothelial IK_Ca and SK_Ca contributes to nitric oxide (NO) generation and is required to elicit endothelium-dependent hyperpolarizations. In the latter responses, the hyperpolarization of the smooth muscle cells is evoked either via electrical coupling through myo-endothelial gap junctions or by potassium ions, which by accumulating in the intercellular space activate the inwardly rectifying potassium channel Kir2.1 and/or the Na⁺/K⁺-ATPase. Additionally, endothelium-derived factors such as cytochrome P450-derived epoxyeicosatrienoic acids and under some circumstances NO, prostacyclin, lipoxygenase products and hydrogen peroxide (H₂O₂) hyperpolarize and relax the underlying smooth muscle cells by activating BK_Ca. In contrast, cytochrome P450-derived 20-hydroxyeicosatetraenoic acid and various endothelium-derived contracting factors inhibit BK_Ca. Aging and cardiovascular diseases are associated with endothelial dysfunctions that can involve a decrease in NO bioavailability, alterations of EDHF-mediated responses and/or enhanced production of endothelium-derived contracting factors. Because potassium channels are involved in these endothelium-dependent responses, activation of endothelial and/or smooth muscle K_Ca could prevent the occurrence of endothelial dysfunction. Therefore, direct activators of these potassium channels or compounds that regulate their activity or their expression may be of some therapeutic interest. Conversely, blockers of IK_Ca may prevent restenosis and that of BK_Ca channels sepsis-dependent hypotension.

Introduction

The vascular endothelium is involved in many different physiological functions, including metabolism, angiogenesis, haemostasis, inflammation, synthesis and degradation of the extracellular matrix as well as the regulation of vascular permeability and vascular tone. The endothelium maintains the balance between vasodilatation and vasoconstriction, inhibition and promotion of the proliferation and migration of smooth muscle cells, prevention and stimulation of the adhesion and aggregation of platelets as well as thrombogenesis and fibrinolysis. Upsetting this tightly regulated balance leads to endothelial dysfunction (Félétou and Vanhoutte, 2006).

The actions of endothelium-derived vasoactive factors in many cases involve the control of membrane potential of the vascular smooth muscle (Félétou and Vanhoutte, 2007a,b). The hyperpolarization of the smooth muscle cells is a powerful mean to produce relaxation. The major effect is a decrease in Ca²⁺ influx by reducing the open probability of voltage-dependent calcium channels (Ca_V; Nelson et al., 1990; Bolton et al., 2002) and the Ca_V-dependent activation of the sarcoplasmic reticulum (Del Valle-Rodriguez et al., 2003). Additionally, hyperpolarization could diminish the release of Ca²⁺ from intracellular stores by decreasing the turnover of intracellular phosphatidylinositol (Itoh et al., 1992). Conversely, the depolarization of vascular smooth muscle cells elicits contraction by opening Ca_V and favouring calcium-induced calcium release. Potassium channels, and especially calcium-activated potassium channels (K_Ca), are key molecules to regulate these membrane electrical events (Nelson and Quayle, 1995). In the endothelial cells, they contribute to the increase in intracellular calcium concentration and therefore regulate the release of endothelium-derived vasoactive factors. Furthermore, they are the key molecular constituents underlying the generation of endothelium-derived hyperpolarizing substances and EDHF-mediated responses (Félétou and Vanhoutte, 2005; Ledoux et al., 2006).

Calcium-activated potassium channels

Potassium channels are the largest and most diverse subgroup of ion channels and are classified in four subgroups according to their membrane topology (Gutman et al., 2003; Alexander et al., 2008). The K_Ca channel family is divided into two subfamilies, the small (SK_Ca) and intermediate (IK_Ca) conductance K_Ca subfamily including K_Ca2.1, K_Ca2.2, K_Ca2.3 (also known as SK1, SK2 and SK3) and K_Ca3.1 (also known as IK1 or SK4) subunits and the large conductance (MaxiK or BK_Ca) K_Ca subfamily including the K_Ca1.1 α subunit (also known as Slo1α).

SK_Ca and IK_Ca channels

Calcium-activated potassium channels of small conductance and IK_Ca channels are voltage-independent and their calcium sensitivity is ascribed to the association with calmodulin (Köhler et al., 1996; Joiner et al., 1997; Xia et al., 1998; Fanger et al., 1999). Tetraethylammonium and tetrabutylammonium are non-specific blockers of K_Ca. SK_Ca channels are specially blocked by the bee toxin apamin and by some scorpion toxins such as scyllatoxin (also named leiurotoxin I from Leiurus quinquestriatus), tamapin (Mesobuthus tamulus) and BmSKTx1 (Buthus martensi, Xu et al., 2004). The plant alkaloid tubocurarine and the synthetic compound UCL-1684 (Campos Rosa et al., 2000) are also potent and reasonably specific blockers of SK_Ca channels (Castle, 1999; Dunn, 1999, Strobaek et al., 2000; Liegeois et al., 2003). Tamapin appears to be selective towards SK2 over SK1 and to a lesser extent over SK3 (Pedarzini et al., 2002).

Non-specific blockers of IK_Ca channels include the scorpion toxins charybdotoxin (Leiurus quinquestriatus) and maurotoxin (Maurus palmatus) as well as clotrimazole, a non-peptide inhibitor of cytochrome P450 monooxygenase. The analogues of clotrimazole, TRAM-34 and TRAM-39, are devoid of cytochrome P450 epoxygenase inhibitory properties and are considered as specific blockers of IK_Ca (Wulff et al., 2000), although the former is also a blocker of non-selective cation channels (Schilling and Eder, 2007).

1-EBIO (1-ethyl-2-benzimidazolinone), its more potent analogue DCEBIO (dichloro-1-ethyl-2-benzimidazolinone), chlorzoxazone-related compounds and riluzole are weak and non-specific activators of IK_Ca and SK_Ca (Cao et al., 2001; Wulff et al., 2007). Another derivative of 1-EBIO, NS-309, is a much more potent opener of both IK_Ca and SK_Ca, with a preferential selectivity for the former (Strøbaek et al., 2004; Leuranguer et al., 2008). CyPPA [cyclohexyl-[2-(3,5-dimethyl-pyrazol-1-yl)-6-methyl-pyrimidin-4-yl]-amine] is a preferential positive modulator of SK2 and SK3 over SK1 but is less potent than NS-309 and possesses inhibitory properties versus BK_Ca and some Na⁺ channels (Hougaard et al., 2007; Wulff et al., 2007).

BK_Ca channels

Calcium-activated potassium channels of large conductance are characterized by a high unitary conductance and are both voltage- and calcium-regulated potassium channels. Numerous isoforms of the Slo1α subunit are generated by alternative splicing. (Meera et al., 2001; Latorre and Brauchi, 2006). In addition, the expression of accessory β subunits (β1 to β4) can lead to channel diversity (Shieh et al., 2000).

Calcium-activated potassium channels of large conductance are also blocked by charybdotoxin and low concentrations of tetraethylammonium. Another scorpion toxin, iberiotoxin (Buthus tamulus), and mycotoxins such as paxilline and penitrem A as well as the synthetic and non-peptide compound 1-[1-hexyl-6-(methyloxy)-1H-indazol-3-yl]-2-methyl-1-propanone (HMIMP; Zeng et al., 2008), are potent and selective inhibitors of this channel. They are activated by synthetic compounds such as NS-11021 the benzimidazolone derivatives, NS-1619 and NS-004, and naturally occurring compounds such as pimaric acid (Gribkoff et al., 2001; Meera et al., 2001; Bentzen et al., 2007; Nardi and Olesen, 2008).

Calcium-activated potassium channels and vascular smooth muscle cells

SK_Ca and IK_Ca channels in vascular smooth muscle cells

In contrast to intestinal smooth muscle, there is little evidence for a functional role of SK_Ca channels in vascular smooth muscle cells, although a non-identified apamin-sensitive and voltage-dependent conductance has been reported (Gebremedhin et al., 1996; Quignard et al., 2000b; Gauthier et al., 2004; 2008). Similarly, in healthy and freshly isolated vascular smooth muscle cells IK_Ca channels are not or very poorly expressed (Figure 1). However, in proliferating cells, as seen in culture or after vascular injury, the expression of this channel increases dramatically (Neylon et al., 1999; Kohler et al., 2003; Tharp et al., 2006; 2008). Up-regulation of IK_Ca is necessary for mitogen-induced suppression of smooth muscle-specific marker genes, that is, differentiation of vascular smooth muscle cells, as well for their proliferation and migration (Tharp et al., 2006). Selective blockade of IK_Ca with TRAM-34 (Wulff et al., 2000) prevents smooth muscle phenotypic changes and coronary artery neointimal formation in two different models of post-angioplasty restenosis (Kohler et al., 2003; Tharp et al., 2008). Similarly, coronary arteriolar remodelling in L-arginine-methylester (L-NAME)-treated rats and post-ischaemic cardiovascular remodelling in rats subjected to coronary artery ligation are associated with an increase in IK_Ca channel expression in vascular smooth muscle cells. A treatment with a statin in the former model and with an AT1 receptor antagonist in the latter prevented both the up-regulation of IK_Ca expression and the structural alterations (Saito et al., 2002; Terata et al., 2003).

Figure 1.

Potassium channels in smooth muscle cells of the guinea pig carotid artery. Effect of the combination of different inhibitors of potassium channels in freshly isolated smooth muscle cells of the guinea pig carotid artery (whole cell configuration of the patch-clamp technique). (A) Large global currents observed in the presence of intracellular calcium for a ramp depolarization from −100 to +80 mV. This current is partially inhibited by iberiotoxin (IbTx), indicating the presence of calcium-activated potassium channels of large conductance (BK_Ca). The addition of charybdotoxin (ChTx) does not produce any further inhibition, suggesting that neither the calcium-activated potassium channels of intermediate conductance (IK_Ca)-dependent current nor the A-type rapidly inactivating transient outward current (K_TO) is activated. The addition of apamin (Apa) produces a further inhibition, indicating the presence of a calcium-activated potassium channels of small conductance ‘(SK_Ca)-like’ current. The subsequent addition of 4-aminopyridine (4-AP) blocked the remaining global outward current demonstrating the contribution of voltage-activated potassium channels (K_V). (B) Large outward current observed for a step depolarization from −100 to +60 mV, confirming the contribution of BK_Ca, ‘SK_Ca-like’ and K_V channels in the recorded current. The effect of 4-AP is not shown for the sake of clarity. Modified from Quignard et al. (Br J Pharmacol 2000b).

BK_Ca channels in vascular smooth muscle cells

Calcium-activated potassium channels of large conductance are expressed in virtually all vascular smooth muscle cells (Figure 1). They are composed of the association of the Slo1α and the β1 subunits. BK_Ca channels, often clustered in groups of 20–100 units, are activated by calcium sparks, localized elemental calcium release events from internal calcium stores and generate spontaneous transient outward currents (STOCs) (Figure 2). These calcium sparks paradoxically lead to a decreased overall intracellular calcium concentration and thus to relaxation of arterial smooth muscle (Nelson et al., 1995; Bychkov et al., 1997; Perez et al., 1999). BK_Ca must be seen as a physiological brake, a feedback inhibitor of the increase in intracellular calcium concentration (Quignard et al., 2000b; 2003; Meera et al., 2001; Ledoux et al., 2006).

Figure 2.

Spatial distribution and functions of calcium-activated potassium channels in the vascular wall. SK_Ca (SK3) and IK_Ca (IK1) are preferentially expressed in ECs. SK_Ca are preferentially located in caveolin-rich domains, at sites of homocellular endothelial gap junctions. A global increase in endothelial [Ca²⁺]_i preferentially activates SK_Ca and an EDHF-mediated response involving a Cx-dependent pathway (Dora et al., 2008). IK_Ca are preferentially localized at the sites of endothelial projections towards the underlying smooth muscle cells. Co-localized sarcoplasmic reticulum elements and associated local calcium release (calcium pulsar) regulate IK1 activation and vascular tone via potassium efflux and subsequent activation of smooth muscle Na⁺/K⁺-ATPase (Ledoux et al., 2008b). The activation of these two channels also favours calcium entry, amplifying NO production (Stankevicius et al., 2006). When BK_Ca channels are expressed in the ECs they can also contribute to this latter mechanism (Brakemeier et al., 2003). BK_Ca are preferentially expressed in smooth muscle cells and are often clustered in large groups. They are activated by calcium sparks or following a global increase in smooth muscle [Ca²⁺]_i. They generate STOCs, leading to arterial smooth muscle hyperpolarization and relaxation (Perez et al., 1999). IK_Ca are expressed in VSMCs undergoing de-differentiation and are involved in their proliferation (Neylon et al., 1999). BK_Ca, calcium-activated potassium channels of large conductance; Cx, connexin; EC, endothelial cell; EDCF, endothelium-derived contracting factor; eNOS, endothelial nitric oxide synthase; IK_Ca, calcium-activated potassium channels of intermediate conductance; IP₃R, IP₃ recepteur; Kir, inwardly rectifying potassium channel; NO, nitric oxide; RyR, ryanodine receptor; SK_Ca, calcium-activated potassium channels of small conductance; STOCs, spontaneous transient outward currents; TRPV4, vanilloid transient receptor potential channel 4; VSMC, vascular smooth muscle cell.

BK_Ca and endothelial vasoactive factors

Endothelium-derived relaxing factors

The relaxations in response to prostacyclin (PGI₂) and its synthetic analogues (beraprost, iloprost, cicaprost) as well as to nitric oxide (NO) and NO donors (nitroglycerin, NONOates, sodium nitroprusside, etc.) are often associated with the concomitant hyperpolarization of the smooth muscle cells. This can involve the opening of ATP-sensitive potassium channels (K_ATP; Parkington et al., 1993; 2004; Corriu et al., 1996; 2001; Quignard et al., 2000a), voltage-activated potassium channels (K_V; Yuan et al., 1996; Li et al., 1997), inwardly rectifying potassium channels (Kir) (Schubert et al., 2004; Orie et al., 2006), two-pore-domain potassium channels (K_2P; Olschewski et al., 2006) and BK_Ca (Robertson et al., 1993; Schubert et al., 1996; Clapp et al., 1998; Quignard et al., 2000a). NO can activate BK_Ca via a protein kinase G-dependent phosphorylation of the channel (Archer et al., 1994) (Figures 3 and 4) or by a direct, cyclic-guanosine monophosphate (cGMP)-independent manner (Bolotina et al., 1994; Mistry and Garland, 1998). In vascular smooth muscle, the cGMP-dependent activation of these channels involves the phosphorylation of either the Slo1α or the β1-regulatory subunits (Nara et al., 2000; Kudlacek et al., 2003) (Figure 4). The cGMP-independent mechanism may involve the binding of NO (or of one of its oxidized derivatives) to thiols, most likely cysteine residues located on the α subunit, in order to form S-nitrosothiols that may establish disulfide bridges with other reduced thiols (Abdelrrahmane et al., 1998; Lang et al., 2000; 2003). NO can also indirectly activate BK_Ca by preventing the formation of an endogenous inhibitor of these channels. By binding to the haem moiety of the cytochrome P450 monooxygenase, NO inhibits the enzymatic formation of 20-Hydroxyeicosatetraenoic acid (20-HETE), a potent inhibitor of BK_Ca activity (Alonso-Galicia et al., 1997; Sun et al., 1998).

Figure 3.

Contribution of BK_Ca in NO-induced relaxation. (A) In rat aorta, acetylcholine-induced endothelium-dependent relaxation is blocked by the guanylate cyclase inhibitor ODQ and partially inhibited by IbTX, indicating that endogeneous NO induces cGMP-dependent relaxation and that the activation of BK_Ca contributes to the mechanism of this relaxation. (B) In rat aorta, a NO donor, DEA-NONOate, also produces cGMP-dependent relaxations, which involves the activation of BK_Ca. (C) In rat aorta, a NO-independent activator of soluble guanylate cyclase, BAY 58-2667, produces a slowly developing relaxation, which involves the activation of BK_Ca. (D) In freshly isolated smooth muscle cells from rabbit carotid artery SIN-1, a NO donor increases the amplitude of the BK_Ca current elicited by a step depolarization (whole cell configuration of the patch-clamp technique, left panel). The effect of SIN-1 is associated with an increase in the open probability of the channel (cell-attached configuration of the patch-clamp technique, right panel; modified from Quignard et al. (Eur J Pharmacol, 2000a). BK_Ca, calcium-activated potassium channels of large conductance; cGMP, cyclic-guanosine monophosphate; IbTX, iberiotoxin; NO, nitric oxide; sGC, soluble guanylate cyclase.

Figure 4.

Endothelial vasoactive factors and smooth muscle BK_Ca activity. 20-HETE, 20-hydroxyeicosatetraenoic acid; BK_Ca, calcium-activated potassium channels of large conductance; cAMP, cyclic-adenosine monophosphate; cGMP, cyclic-guanosine monophosphate; COX, cyclooxygenase; EC, endothelial cell; EDCFs, endothelium-derived contracting factors; EDRFs, endothelium-derived relaxing factors; EETs, epoxyeicosatrienoic acids; eNOS, endothelial nitric oxide synthase; GS, G-protein S; GSα, α subunit of G-protein S; H₂O₂, hydrogen peroxide; IP, prostcyclin receptor; NO, nitric oxide; NOX, NADPH oxidase; P450 2C-2J or 4A, cytochrome P450 monooxygenase 2C, 2J or 4A; PGI₂, prostacyclin; R?, putative receptor; ROS, reactive oxygen species; sGC, soluble guanylate cyclase; SOD, superoxide dismutase; TP, thromboxane/endoperoxide receptor; TRPV4, vanilloid transient receptor potential channel 4; VSMC, vascular smooth muscle cell; XO, xanthine oxidase.

The activation of BK_Ca is the preponderant mechanism of the relaxation produced by epoxyeicosatrienoic acids (EETs), generated by endothelial cytochrome P450 epoxygenases (Campbell et al., 1996; Popp et al., 1996; Fisslthaler et al., 1999; Gauthier et al., 2005; Huang et al., 2005; Weston et al., 2005). They do not directly activate these channels as shown for other fatty acids (Ordway et al., 1989). EETs may interact with ‘receptor(s)’, which remain to be identified, and promote the phosphorylation and the activation of the Slo1α subunit of BK_Ca (Li and Campbell, 1997; Li et al., 1999). EETs can also activate BK_Ca in much more indirect manners via the generation of carbon monoxide (CO) (Sacerdoti et al., 2006) or the activation of the vanilloid transient receptor potential channel 4 (TRPV4) and the subsequent induction of STOCs (Earley et al., 2005) (Figure 4).

Hydrogen peroxide (H₂O₂), depending on the blood vessel, the presence of the endothelium, the experimental conditions or the concentrations studied, possesses dilator or constrictor properties. The production of H₂O₂ can be involved in endothelium-dependent relaxations in response to agonists and flow (Miura et al., 2003; Hatoum et al., 2005) or compensate for the decreased production of NO (Cosentino and Katusic, 1995). This reactive oxygen species could play an important role in coronary auto-regulation (Yada et al., 2003) and a cardioprotective role during ischaemia-reperfusion injury (Yada et al., 2006). H₂O₂-induced relaxations involve multiple pathways, including the hyperpolarization of the vascular smooth muscle cells by activating BK_Ca, either directly or following soluble guanylyl cyclase stimulation (Hayabuchi et al., 1998; Thengchaisri and Kuo, 2003) (Figure 4). However, H₂O₂ can also be a potent inhibitor of BK_Ca by altering cysteins on the Slo1 α subunit (Tang et al., 2004).

Endothelium-derived contracting factors

Potassium channels also modulate the action of endothelium-derived contracting factors (EDCFs). For instance, depending on the blood vessels, endothelin-1 inhibits various populations of potassium channels, Kir (Park et al., 2005a), K_ATP (Miyoshi et al., 1992; Park et al., 2005b), K_V (Shimoda et al., 2001) and BK_Ca (Minami et al., 1995), which contribute to its depolarizing effect. Similarly, the activation of the thromboxane/endoperoxide TP receptor inhibits K_V (Cogolludo et al., 2003), BK_Ca (Scornik and Toro, 1992) and depolarizes the smooth muscle cells (Corriu et al., 2001). Reactive oxygen species can directly contribute to the contraction of the vascular smooth muscle. They increase the sensibility of the contractile proteins to calcium ions (Jin et al., 1991) and depolarize the vascular smooth muscle cells by inhibiting the activation of K_ATP (Kinoshita et al., 2004), K_V (Liu et al., 2001; Li et al., 2004) and BK_Ca (Brzezinska et al., 2000; Liu et al., 2002a; Tang et al., 2004) (Figure 4).

20-Hydroxyeicosatetraenoic acid, a metabolite of the cytochrome P450 4A or 4F family, is a potent endogenous vasoconstrictor of renal, cerebral, coronary, mesenteric and skeletal muscle arteries (Miyata and Roman, 2005). The vasoconstrictor effect of 20-HETE can involve endothelium-dependent effects such as the cyclooxygenase (COX)-dependent production of thromboxane A₂ (Randriamboavonjy et al., 2003) or a direct effect on the smooth muscle cells. In this latter case, the main target of 20-HETE is the inhibition of BK_Ca and the subsequent activation of Ca_V, thus depolarizing and contracting the vascular smooth muscle cells (Ma et al., 1993; Harder et al., 1994; Zou et al., 1996; Gebremedhin et al., 1998; Sun et al., 1999; Obara et al., 2002) (Figure 4). Cytochrome P450 4A overexpression in the blood vessel wall increases the production of 20-HETE and causes hypertension and endothelial dysfunction (Wang et al., 2006). Conversely, specific inhibition of the synthesis of 20-HETE reduces infarct size after ischaemic stroke and reverses the delayed vasospasm in models of subarachnoid haemorrhage (Miyata et al., 2005; Takeuchi et al., 2005; Hacein-Bey et al., 2006). The formation of 20-HETE and the expression of cytochrome P450 enzymes are altered in many animal models of cardiovascular diseases and in some forms of human hypertension (Sarkis and Roman, 2004).

BK_Ca and genetically modified animals

In mice with a disrupted gene for the auxiliary β1 subunit of BK_Ca, the generation of calcium sparks in vascular smooth muscle cells are of normal amplitude and frequency, but the frequency of STOCs is reduced. The contractile responses of isolated aortic rings in the knockout mice are increased when compared with aortas from wild type controls, and the systemic arterial blood pressure is higher in the former than in the latter (Brenner et al., 2000; Pluger et al., 2000). Furthermore, the deletion of the β1 subunit causes smooth muscle depolarization, a subsequent increase in NADPH oxidase-dependent production of superoxide anion and endothelial dysfunction due to reduced cGMP-dependent kinase-I activity (Oelze et al., 2006). Depolarization is also an important stimulus for endothelial superoxide generation (Sohn et al., 2000).

Mice knockout for the Slo1 α subunit exhibit a moderate increase in blood pressure attributed to vascular abnormalities (absence of STOC and decrease in the effectiveness of cGMP/cGMP kinase pathway) as well as primary hyperaldosteronism (Sausbier et al., 2005). These mice have marked erectile dysfunction and are less responsive to the phosphodiesterase-5 (PDE-5) inhibitor sildenafil (Werner et al., 2005; 2008).

BK_Ca and cardiovascular diseases

In smooth muscle cells from several arteries of various models of hypertension, larger BK_Ca currents are observed than in those of normotensive controls (Cox, 2002; Eichhorn and Dobrev, 2007). Furthermore, because BK_Ca are less sensitive to the deleterious effects of reactive oxygen species than other potassium channels, they could play a compensatory role in diabetes and atherosclerosis (Liu and Gutterman, 2002). A compensatory overexpression of the β1 subunit has been reported in aging rats and of the α subunit in spontaneously hypertensive rats (SHR; Liu et al., 1998; Nishimaru et al., 2004a; Chang et al., 2006). In HEK293 cells transfected with the two α and β1 subunits of the human BK_Ca channel, hypoxia induces also an increased expression of the β1 subunit (Hartness et al., 2003). In SHR cerebral arteries, the function and expression of BK_Ca are not altered (Nishimaru et al., 2004a).

Nevertheless, peroxynitrite and H₂O₂ can inhibit BK_Ca activity as shown in human coronary artery smooth muscle cells (Liu et al., 2002b; Lu et al., 2006). A reduced activity of BK_Ca channels has also been reported in arteries from various animal models. This could be associated with no change in the expression of the β1 subunit as in the Zucker diabetic rat (Burnham et al., 2006a) or with a down-regulation of the expression of the β1 subunit as in SHR arteries (Amberg and Santana, 2003), in angiotensin II-dependent hypertensive rats (Amberg et al., 2003), in streptozotocin-treated rats and mice (McGahon et al., 2007; Dong et al., 2008) and in hypoxic rats or human mammary artery subjected to hypoxia (Navarro-Antolín et al., 2005). Alternatively, the expression of the Slo1α subunit is decreased in rat with pulmonary hypertension (Bonnet et al., 2003), in L-nitro-arginine-hypertensive rats (Bratz et al., 2005a,b), and the expression of both subunits is decreased in aging rats (Marijic et al., 2001; Nishimaru et al., 2004b). Vascular bed-specific alterations in BK_Ca channel expression/activity are likely to explain these discrepancies.

A genetic polymorphism in the β1 subunit of the human BK_Ca channel (E65K), a gain-of-function mutation, is linked to a reduced diastolic hypertension in aging woman and is one of the strongest genetic factors associated so far to protection against infarction and stroke (Sentíet al., 2005).

BK_Ca and cardiovascular therapeutic opportunities

Although significant evidence suggests that BK_Ca channels play a crucial role in many patho-physiological conditions, including cardiovascular diseases such as hypertension, ischaemic heart disease, stroke and erectile dysfunction, at the present time all clinical trials involving BK_Ca openers, at the exception of a single one for the potential treatment of bronchial asthma and chronic obstructive pulmonary diseases, have been discontinued (Nardi and Olesen, 2008). The poor potency and the lack of selectivity of the early compounds may explain the failures of these multiple clinical trials. Nevertheless, the mechanism of action underlying the beneficial effects of some cardiovascular drugs currently prescribed can involve the activation of BK_Ca channels, for instance NO donors (Khan et al., 1998; Gruhn et al., 2002), PDE inhibitors such as sildenafil (Gragasin et al., 2004; Werner et al., 2008), or cilostazol (Park et al., 2006), PGI₂ analogues (Clapp et al., 1998; Tanaka et al., 2004), hydralazine (Bang et al., 1998), acetazolamide (Pickkers et al., 2001) and xenoestrogens (Dick and Sanders, 2001; Nardi and Olesen, 2008). Alternatively, some drugs such as dehydroepiandrosterone increase their expression (Bonnet et al., 2003).

In contrast, some drugs can inhibit BK_Ca channels, for instance verapamil (Harper et al., 2001), or simvastatin (Seto et al., 2007), although the so-called pleitropic effect of statins, by up-regulating and activating endothelial NO synthase (NOS; Rikitake and Liao, 2005) may counterbalance the direct inhibitory effect of simvastatin on BK_Ca channel activation.

In patients with severe erectile dysfunction, the first small scale clinical trial ever performed with the gene transfer of the hSlo1α subunit has provided promising results (Melman et al., 2006). In the other hand, inhibiting the activation of BK_Ca channels might prove beneficial in the hypotensive state associated with shock (Zhao et al., 2007).

Calcium-activated potassium channels in vascular endothelial cells

BK_Ca channels in endothelial cells

In most endothelial cells, when freshly isolated, BK_Ca channels are poorly expressed and iberiotoxin-sensitive currents are observed only at very positive potentials (Marchenko and Sage, 1996; Kohler et al., 2000; Bychkov et al., 2002; Gauthier et al., 2002; Ledoux et al., 2008a) (Figure 5). This can possibly be attributed to the absence in these cells of regulatory BK_Caβ subunits that enhance the Ca²⁺ sensitivity (Rusko et al., 1992; Papassotiriou et al., 2000). However, when expressed in endothelial cells, BK_Ca channels regulate NO production (Brakemeier et al., 2003).

Figure 5.

Potassium currents in porcine coronary arterial endothelial cells. Effect of the combination of different inhibitors of potassium channels in freshly isolated porcine coronary arterial endothelial cells (whole cell configuration of the patch-clamp technique). (A) Representative K⁺-currents elicited by 10 mV voltage steps in control, after application of iberiotoxin (IbTX), iberiotoxin + apamin (IbTX + Apamin) and iberiotoxin + apamin + charybdotoxin (IbTX + Apamin + ChTX). (B) Summary bar graph: the presence of apamin (Apa.) and charybdotoxin produced a statistically significant inhibition of the amplitude of the K⁺-currents.

SK_Ca and IK_Ca channels in endothelial cells

In contrast, the IK_Ca and SK_Ca channels, especially the SK3 α subunit, are constitutively expressed in endothelial cells (Marchenko and Sage, 1996; Kohler et al., 2000; 2001; Burnham et al., 2002; Bychkov et al., 2002) (Figure 5). IK_Ca and SK_Ca channels have a different spatial distribution in endothelial cells. In the rat mesenteric artery, SK_Ca are preferentially located at sites of homocellular endothelial gap junctions and caveolin-rich domains and are associated with various connexins (Cx) while IK_Ca are preferentially localized at the sites of endothelial projections often associated with myo-endothelial gap junctions (Sandow et al., 2006; Absi et al., 2007; Dora et al., 2008; Ledoux et al., 2008b) (Figure 2).

SK_Ca, IK_Ca and endothelial function

Agonists that stimulate G protein-coupled receptors as well as calcium ionophores, thapsigargin and cyclopiazonic acid increase the endothelial intracellular calcium Ca²⁺ concentration, which activates these two potassium channels. This leads to the hyperpolarization of the endothelial cells and evoke endothelium-dependent hyperpolarizations of vascular smooth muscle cells. The hyperpolarization of the endothelial cells in turn favours the entry of calcium by increasing the driving force for this ion (Busse et al., 1988; Johns et al., 1988; Luckhoff and Busse, 1990a,b; Cheung and Chen, 1992; Fukao et al., 1995; Kamouchi et al., 1999). Therefore, endothelial K_Ca are key players in EDHF-mediated responses (Figure 6) and contribute to the activation of calcium-sensitive enzyme such as the NOS and the generation of NO (Stankevicius et al., 2006; Sheng and Braun, 2007). The endothelium-dependent hyperpolarizations of vascular smooth muscle cells can be evoked by direct electrical coupling through myo-endothelial gap junctions and/or the accumulation of potassium ions in the intercellular space. Potassium ions hyperpolarize the smooth muscle cells by activating Kir2.1 and/or Na⁺/K⁺-ATPase (Edwards et al., 1998; 2003; Griffith et al., 2004; Félétou and Vanhoutte, 2007c) (Figure 2).

Figure 6.

Involvement of SK_Ca and IK_Ca in EDHF-mediated responses. Endothelium-dependent hyperpolarizations to ACh in vascular smooth muscle of the guinea pig carotid artery (in the presence of inhibitors of NOS and COX). (A) Original membrane potential recordings showing the concentration-dependent hyperpolarizing effect of ACh. (B) The concentration- and endothelium-dependent hyperpolarizations produced by ACh are not affected by the presence of ChTx, partially inhibited by that of Apa. and virtually abolished by the combination of the two toxins. (C) Similarly, TRAM-34, a non-peptidic and selective blocker of IK_Ca, does not affect the hyperpolarization elicited by ACh while UCL 1684, a selective blocker of SK_Ca, produces a partial inhibition. The combination of the two blockers prevents the hyperpolarizing effect of ACh. Modified from Gluais et al. (Br J Pharmacol, 2005). ACh, acetylcholine; Apa., apamin; ChTx, charybdotoxin; COX, cyclooxygenase; EDHF, endothelium-derived hyperpolarizing factor; IK_Ca, calcium-activated potassium channels of intermediate conductance; NOS, nitric oxide synthase; SK_Ca, calcium-activated potassium channels of small conductance.

In rats, the number of myo-endothelial gap junctions and the expression of endothelial SK3 and IK1 increase with a reduction in the size of the artery (Sandow and Hill, 2000; Hilgers et al., 2006), a phenomenon that parallels the enhanced contribution of EDHF-mediated responses in endothelium-dependent relaxations (Hwa et al., 1994; Shimokawa et al., 1996). However, in mice this inverse relationship between the size of an artery and an increase in the number of myo-endothelial gap junctions and in the expression of endothelial SK3 and IK1 has not been observed (Ceroni et al., 2007). Nevertheless, in both species, the presence of heterocellular myo-endothelial gap junctions is associated with EDHF-mediated responses (Sandow et al., 2002; Dora et al., 2003). In rat mesenteric artery, antibodies against Cx-40, when loaded selectively into the endothelial cells, show that Cx-40 is localized to the end of endothelial cell projections at myo-endothelial gap junctions. These antibodies block EDHF-mediated responses, demonstrating a critical role for myo-endothelial gap junctions containing Cx-40 in EDHF-mediated dilatation (Mather et al., 2005).

Few experiments involve the in vivo administration of blockers of SK_Ca and IK_Ca channels, in the presence or not of inhibitors of NOS and COX. In anaesthetized rats, intravenous bolus injection of the combination of charybdotoxin plus apamin does not affect basal arterial blood pressure but attenuates the decrease in blood pressure produced by either acetylcholine or ghrelin infusion (Shinde et al., 2005; Desai et al., 2006). Similarly, the acute administration of the two toxins in the rat mesenteric and hindlimb vascular beds does not affect the basal conductance but partially inhibits the effects of acetylcholine and bradykinin (Parkington et al., 2002; Dabisch et al., 2004). These results suggest that EDHF-mediated responses play a physiological role in the regulation of vascular resistance.

EDHF-mediated responses and genetically modified animals

More conclusive evidence has been obtained from genetically modified animals. In NOS-3 knockout, EDHF-mediated responses play a compensatory role for the absence of endothelial NO (Brandes et al., 2000; Ding et al., 2000). The adaptation to NOS deletion is gender-specific (Wu et al., 2001). This gender difference in double knockout mice for NOS-3 and COX-1 involves EDHF-mediated responses. In isolated resistance arteries from double knockout female mice, endothelium-dependent relaxations are preserved and involve exclusively K_Ca channels while in arteries from male ones the endothelium-dependent relaxations are impaired severely. Similarly, bradykinin produces dose-dependent hypotension in female but no effect in double knockout male mice. In female mice, this double deletion does not affect mean arterial blood pressure while the corresponding male are hypertensive (Scotland et al., 2005).

Transgenic mice harbouring genetically targeted alleles for the SK3 channel have been engineered, in which SK3 gene expression can be experimentally controlled by dietary doxycycline (Bond et al., 2000). In those transgenic mice, the level of expression of SK3 channels on the endothelial cells correlates with the cell membrane potential of both endothelial and vascular smooth muscle cells, with the tone of isolated mesenteric arteries and the diameter of these arteries in situ, as well as with the arterial blood pressure of the animals (Taylor et al., 2003). Disruption of the IK1 gene reduces the hyperpolarization of the endothelial and smooth muscle cells in response to acetylcholine. This results in decreased dilatation in the carotid artery and in resistance vessels because of a substantial reduction of EDHF-mediated responses. Moreover, the IK1 deletion significantly increases arterial blood pressure and causes mild left ventricular hypertrophy. These results indicate that in mice, the endothelial IK_Ca is a fundamental determinant of endothelial hyperpolarization and EDHF signalling and, thereby, a crucial determinant in the control of vascular tone and overall circulatory regulation (Si et al., 2006). In double knockout mice, lacking both SK3 and IK1, the addition of the detrimental effects provoked by the deletion of either gene is observed (Brähler et al., 2008; De Wit, 2008). In mice that lack the Kir2.1, but not in those deleted for Kir2.2 genes, Kir currents are absent and stimulation with moderate increases in potassium concentration does not produce relaxation, indicating that the Kir2.1 gene is required for Kir currents and potassium-induced dilatation (Zaritsky et al., 2000). Deletion of TREK-1 in mice leads to an important alteration in vasodilatation of mesenteric arteries induced by acetylcholine and bradykinin. However, in non-pathological animals this channel is associated with NO release but not with EDHF-mediated responses (Garry et al., 2007).

Connexin-37 and Cx-40 are the predominant gap junction proteins in the endothelial cells of the mouse (Simon and McWhorter, 2003). Cx-40 proteins are involved in endothelial homocellular gap junctions and also in heterocellular gap junctions linking endothelial cells not only to smooth muscle cells but also to renin-producing juxtaglomerular cells. The presence of the latter gap junction communication is required in order to maintain the calcium-dependent inhibitor effects of angiotensin II and that of intra-renal pressure on renin secretion and synthesis, suggesting that the endothelium is strongly involved in the regulation of the renin system. Mice deficient for Cx-40 are hypertensive. However, the control of renin production only partially explained the hypertension in Cx-40 knockout mice (Wagner et al., 2007). The arterioles of these animals also exhibit a reduced spread of dilatation in response to endothelium-dependent vasodilators and irregular arteriolar vasomotion (De Wit et al., 2000; 2003; Simon and McWhorter, 2002; Figueroa et al., 2003). Mice subjected to specific deletion of endothelial Cx-43 do not present major alterations in arterial blood pressure (Liao et al., 2001; Theis et al., 2001), possibly because this Cx is not the major one expressed in murine endothelial cells.

Theses results show that deletion of each key molecular component of EDHF-mediated responses is associated with alterations in arterial blood pressure suggesting that this endothelial pathway contributes to the overall regulation of cardiovascular function (Kohler and Hoyer, 2007).

EDHF-mediated responses and endothelial dysfunction

Endothelial dysfunction is observed in various cardiovascular diseases and is often associated with a decrease of NO synthesis and/or a loss of its biological activities. However, alteration of the EDHF pathway can also contribute to these endothelial dysfunctions or conversely can compensate the loss of NO bioavailability (Félétou and Vanhoutte, 2005; 2007c). The alteration of EDHF-mediated responses has been reported with aging and various pathological conditions (hypertension, atherosclerosis, hypercholesterolemia, heart failure, ischaemia-reperfusion, angioplasty, eclampsia, diabetes, sepsis; Félétou and Vanhoutte, 2004; 2005). However, depending on the model or the vascular bed studied, marked differences can be observed.

Hypertension

Hypertension per se does not produce a consistent depression of the EDHF-mediated responses. For instance, in the mesenteric artery of the SHR, the impairment of the endothelium-dependent relaxations is attributed to a marked attenuation of the EDHF component and a concomitant production of COX-derived contractile prostanoids (EDCFs) with no or little alteration in the production of NO, whereas, in the model of L-NAME-induced hypertension and in the same vascular bed, an increase in EDHF-mediated responses may compensate the inhibition of NO production (Félétou and Vanhoutte, 2005). In human with essential hypertension, the mechanism underlying the endothelial dysfunction is also linked to oxidative stress and the activation of COX, which reduces availability of NO. In the forearm vascular bed, the presence of an EDHF-like pathway, possibly a cytochrome P450-dependent mechanism, compensates the decreased NO bioavailability in order to sustain endothelium-dependent vasodilatation (Taddei et al., 2001; Passauer et al., 2003). This compensatory pathway can be depressed by additional aggravating factors such as hyperhomocysteinemia (Taddei et al., 2001). In myometrial arterioles from pre-eclamptic mothers, the up-regulation of the EDHF-mediated responses observed in normal pregnancy does not occur (Kenny et al., 2002). The question remains, whether the impairment of EDHF-mediated responses contributes to the genesis of the syndrome or is a consequence of the hypertensive process.

Hypercholesterolemia-atherosclerosis

Hypercholesterolemia is generally associated with a preserved or an enhanced contribution of EDHF-mediated responses that compensate for the decrease in NO-mediated relaxation (Selemidis and Cocks, 2002). The resistance of endothelium-dependent hyperpolarizations to hypercholesterolemia has been demonstrated in arteries from rabbit (Brandes et al., 1997), SHR (Kagota et al., 1999), APOE-deficient mice (Ding et al., 2005; Morikawa et al., 2005; Wolfle and de Wit, 2005) and dyslipidemic hApoB+/+ mice (Krummen et al., 2005). However, in isolated gastroepiploic arteries from atherosclerotic patients, endothelium-dependent hyperpolarizations are inhibited (Urakami-Harasawa et al., 1997). The prolonged duration of hypercholesterolemia and the severity of the atherosclerotic process in the human may contribute to the degree of dysfunction of the EDHF pathway.

Diabetes

Conversely, at the exception of some murine models, EDHF-mediated responses are depressed in various models of type I and type II diabetes (Félétou and Vanhoutte, 2004). In patients with type I diabetes under good glycemic control and without albuminuria, endothelial function appears normal, and both the NO- and the EDHF-mediated responses are preserved. However, in patients with microalbuminuria, impairment of the endothelium-dependent vasodilatation is observed. In these patients and in patients with type II diabetes, whether or not the various components of the endothelium-dependent vasodilatation are differentially affected by the disease is not yet known (De Vriese et al., 2000).

Alteration of EDHF-mediated responses and changes in SK3 and IK1 expression

An impaired EDHF-mediated response is associated with a decrease expression of endothelial SK3 and/or IK1 channels in carotid arteries from rat subjected to balloon injury (Kohler et al., 2001), and in those of 5/6-nephrectomized rats (Kohler et al., 2005) as well as in mesenteric arteries of ovariectomized rats (Liu et al., 2002a) and of diabetic apo-E−/− mice (Ding et al., 2005). The endogenous inhibitor of NOS, assymetric dimethyl-L-arginine (ADMA), diminishes SK3 expression and its presence is associated with decreased EDHF-mediated responses (Li et al., 2007). In contrast, the reduced EDHF-contribution in mesenteric arteries of Zucker diabetic rats is not associated with a clear change in the pattern of expression of these two endothelial channels. The expression of SK3 is increased but the activity of the channel is markedly reduced while the IK1 expression is slightly reduced without any loss in activity (Burnham et al., 2006b; Weston et al., 2008).

The endothelial dysfunction in mesenteric arteries from db/db mice does not involve an impairment of the EDHF-mediated response, and no change in the expression of SK3 and IK1 is observed (Pannirselvam et al., 2006). However, in mesenteric arteries of angiotensin II-dependent hypertensive rats, EDHF-mediated responses are not compromised although a reduced functional activity and expression of SK3 channels is observed, possibly because the functional activity of IK1 channels compensates for the impaired SK3 activity (Hilgers and Webb, 2007).

In pulmonary artery from rats subjected to monocrotaline-induced pulmonary hypertension, an enhanced expression of SK3 and IK1 is observed along with an augmented EDHF-mediated response (Morio et al., 2007). However, the compensatory increase in EDHF-mediated response observed in endothelial NOS (eNOS)−/− mice is not associated with any concomitant changes in the expression of these two channels (Ceroni et al., 2007).

Therefore, in many instances the alterations in EDHF-mediated responses do not correlate with changes in the expression of endothelial K_Ca channels.

EDHF-mediated responses and therapeutic interventions

Therapeutic interventions with beneficial effects on the cardiovascular system such as angiotensin converting enzyme inhibitors, antagonists of angiotensin receptors and PDE-3 inhibitors (Matsumoto et al., 2005) can restore EDHF-mediated responses, suggesting that the improvement of the EDHF pathway contributes to the observed beneficial effect. Similarly, various so-called non-pharmacological therapeutic strategies including exercise and supplementation with estrogens, omega-3 polyunsaturated fatty acids, polyphenol derivatives, potassium and/or calcium help to reverse endothelial dysfunction including blunted EDHF-mediated responses. Whether or not the improvement of these EDHF-mediated responses contributes to the beneficial effects of these dietary manoeuvres can be suspected but is far from being demonstrated (Félétou and Vanhoutte, 2004; 2005).

EDHF-mediated responses and future therapeutics

The improvement or restoration of EDHF responses has not been, yet, the direct purpose of any pharmaceutical effort. Activating or increasing the expression of endothelial TRP, IK_Ca and/or SK_Ca channels and smooth muscle Kir and/or specific isoform(s) of Na⁺/K⁺-ATPase as well as facilitating myo-endothelial communication and increasing the expression of appropriate Cx (Cx-40, Cx-43 and Cx-37) may represent new potential targets. However, most of these targets are ubiquitously expressed, and proper selectivity might be difficult to achieve. For instance, activating K_Ca channels may appear as a good strategy to improve endothelial function, by enhancing NO release (Stankevicius et al., 2008) and increasing endothelium-dependent hyperpolarizations (Leuranguer et al., 2008) (Figure 7), but IK_Ca channels are required for phenotyping changes in vascular smooth muscle (Neylon et al., 1999; Kohler et al., 2003; Tharp et al., 2006; 2008) and are also involved in the proliferation of endothelial (Grgic et al., 2005) and various cancerous cells (Jäger et al., 2004; Wang et al., 2007). Therefore, activators of IK_Ca may have some unwanted detrimental effects. Furthermore, the precise role of these various targets is far from being completely understood. For instance, a specific and potent agonist of TRPV4, a potentially promising target in cardiovascular diseases since arterial responses to shear stress critically depend on the activation of this endothelial channel (Nilius and Voets, 2004; Hartmannsgruber et al., 2007), has been recently identified (GSK1016790A, Thorneloe et al., 2008). As expected, GSK1016790A increases endothelial intracellular calcium concentration and produces endothelium-dependent relaxations, but also causes endothelial failure, circulatory collapse and death (Willette et al., 2008).

Figure 7.

NS-309 induces endothelium-dependent hyperpolarization. (A) Chemical structure of NS-309, an activator of calcium-activated potassium channels of intermediate and small conductance. (B) Original membrane potential recordings showing the effects of acetylcholine (ACh) and NS-309 in vascular smooth muscle of the guinea pig carotid artery with endothelium (in the presence of inhibitors of nitric oxide synthase and cyclooxygenase). (C) Concentration-dependent effects of NS-309. (D) Summary of the effects of endothelial removal on ACh-, NS-309- and levcromakalim (Levcrom.)-induced hyperpolarization. (E) Effects of the combinations of apamin plus charybdotoxin on ACh-, NS-309- and levcromakalim-induced hyperpolarizations. Modified from Leuranguer et al. (Naunyn Schmiedebergs Arch Pharmacol, 2008).

Taken together, these observations indicate that better (i.e. more potent, more specific and if possible orally active) pharmacological tools must be developed to better understand the role of the various molecular constituents underlying EDHF-mediated responses. Then, it shall be possible to determine whether or not putative cardiovascular targets identified within this pathway are drugable.

Glossary

Abbreviations:

1-EBIO

1-ethyl-2-benzimidazolinone

20-HETE

20-hydroxyeicosatetraenoic acid

ADMA

assymetric dimethyl-L-arginine

BK_Ca

calcium-activated potassium channels of large conductance

Ca_V

voltage-dependent calcium channels

cGMP

cyclic-guanosine monophosphate

CO

carbon monoxide

COX

cyclooxygenase

Cx

connexin

CyPPA

cyclohexyl-[2-(3,5-dimethyl-pyrazol-1-yl)-6-methyl-pyrimidin-4-yl]-amine

DCEBIO

dichloro-1-ethyl-2-benzimidazolinone

EDCF

endothelium-derived contracting factor

EDHF

endothelium-derived hyperpolarizing factor

EETs

epoxyeicosatrienoic acids

H₂O₂

hydrogen peroxide

IK_Ca

calcium-activated potassium channels of intermediate conductance

K_2P

two-pore-domain potassium channels

K_ATP

ATP-sensitive potassium channels

K_Ca

calcium-activated potassium channels

Kir

inwardly rectifying potassium channels

K_V

voltage-activated potassium channels

L-NAME

L-arginine-methylester

NO

nitric oxide

NOS

nitric oxide synthase

PDE

phosphodiesterase

PGI₂

prostacyclin

SHR

spontaneously hypertensive rats

SK_Ca

calcium-activated potassium channels of small conductance

STOCs

spontaneous transient outward currents

TRPV4

vanilloid transient receptor potential channel 4

Conflict of interest

Employee of the Private Pharmaceutical Company Servier.