Keywords

Introduction

Ca 2+ -activated K + channels (K Ca ) are familiar to most electrophysiologists because KCa channels produce afterhyperpolarizations following an action potential (AP) in neurons and thereby influence firing pattern and neurotransmission [1]. Cardiovascular researchers recognize KCa as important effector proteins in endothelial control of arterial tone [2–5]. This is based on observations that subtype-selective inhibition of the major vascular KCa, K Ca 3.1, K Ca 2.3, and K Ca 1.1 impair the endothelium-derived hyperpolarization factor (EDHF)- type vasodilation (involving smooth muscle KCa1.1 as target of a diffusible factor) and the endothelium-derived hyperpolarization (EDH)- type (involving endothelial KCa3.1/KCa2.3 as initiators and gap-junctional coupling of endothelium and smooth muscle) (Fig. 5.1). Both mechanisms have in common that they converge on hyperpolarization of smooth muscle leading to closure of voltage-gated Ca2+ channels and the resulting drop in intracellular Ca2+ causes finally myocyte relaxation and vasodilation. Hence, EDH or EDHF pathways are mechanistically different from pathways involving other diffusible molecules like NO or prostaglandins, although the latter paracrine factors are also capable in some cases to induce smooth muscle hyperpolarization and thus to act as EDHFs and endothelial hyperpolarization stimulates Ca2+-dependent NO-synthesis by enhancing Ca2+ dynamics after receptor stimulations.

Fig. 5.1
figure 1

Schematic overview of putative mechanisms of endothelium-derived-hyperpolarization mediated dilation(EDH) and endothelium-derived hyperpolarizing Factor (EDHF)-mediated dilation. CAM calmodulin, eNOS endothelial nitric oxide synthase, MEGJ myo-endothelial gap junction, STOCs spontaneous transient outward currents

Full size image

At present, the relative physiological significance of either EDHF or EDH systems is still debated, when discussing the input of the respective systems to human vasculature and in cardiovascular disease [4]. Yet, it is clear that the EDHF- or EDH-triggered hyperpolarization of smooth muscle is a potent mechanism that can stand on its own. However, it is not fully understood when and under what conditions EDHF and EDH mechanisms contribute alone, in parallel, or co-operate (or not) with NO-signaling or other diffusible dilating or contracting factors to vasoregulation in the different vascular beds. But it is likely that they integrate in a very complex fashion with NO being without any doubt a major player (for recent extensive review see [6]).

Here, we review mainly the contributions of the authors to their specific field according to the Editors’ wish and start with a short introduction into KCa genes and their expression in the vascular wall, defects of EDH signaling and blood pressure alterations caused by KCa3.1/KCa2.3 deficiency, followed by KCa-pharmacology. In the following, the main focus will be on the cardiovascular actions of novel small molecule modulators (activators and inhibitors) targeting endothelial KCa3.1 and KCa2.3 channels in arteries and systemically. Finally, we explore potential therapeutic possibilities and briefly elaborate on possible impacts on cardiac and neurological functions. For additional in-depth and complete reviews and to not diminish previous work by others and us, we wish to re-direct the reader to [4, 7, 8].

KCa Genes, General Biophysical Properties of KCa, and Genetic Variations Relevant for Cardiovascular Disease

The gene family of KCa channels is divided into two subgroups according to phylogenetic relationships, small or large unitary conductance, and the mechanisms of Ca2+ activation [9] (Fig. 5.2a). The first group comprises four subtypes, of which three have a small unitary conductance of 5–10 pS (gene names: KCNN1, KCNN2, and KCNN3 , protein names SK1-SK3, or K Ca 2.1-K Ca 2.3) and one member with an intermediate conductance of 20–40 pS (gene name: KCNN4; protein name: IK, IKCa, or K Ca 3.1). The functional channels are tetramers formed by four subunits and the four pore loops located between S5 and S6 of each subunit form the ion conduction pathway. Ca2+ sensitivity is conferred by four calmodulin molecules constitutively bound to the four CAM-bindings domains located in the cytosolic c-terminus.

Fig. 5.2
figure 2

(a) Phylogenetic relationships of Ca2+-activated K+ channel genes. (b) Schematic representation of KCa currents and their sensitivity to peptide toxins. Upper panel: Illustration of voltage-independent Ca2+-activated KCa2.3/KCa3.1 currents in endothelial cells and inhibition by the peptide toxins, apamin (Apa) and charybdotoxin (ChTx). U Rest resting membrane potential; −89 mV is the K equilibrium potential. Note that the KCa2.3/KCa3.1 channel activation produces voltage-independent currents shifting the reversal potential towards the K equilibrium potential. Lower panel: Ca2+ activation of voltage-dependent KCa1.1 currents and inhibition by iberiotoxin (IbTx) or ChTx. Note that dual Ca2+ and voltage activation occurs at near the URest, stabilizing the membrane potential

Full size image

An important feature to understand the roles of these channels in cardiovascular biology is that the channels lack a voltage sensor and are thus voltage-independent, which means that the channels do not inactivate at negative membrane potential like e.g. voltage-gated KV channels. This feature enables KCa2 and KCa3.1 channels to produce long-standing K+-outward currents at potentials more negative than the resting potentialFootnote 1 and by this they produce persisting membrane hyperpolarization (e.g. to −50 up to max. −89 mV (K+ equilibrium potential) (Fig. 5.2b). In vascular biology, such a durable hyperpolarization is required for complete closure of smooth muscle voltage-gated Ca2+ channels and thereby for the endothelium-dependent hyperpolarization (EDH)- type of arterial dilation [8].

In other tissues, KCa2 (and possibly KCa3.1) produce afterhyperpolarization in neurons and thereby regulate firing frequency [1] and—in the heart—co-determine the length of the atrial action potential and thus atrial repolarization and the contraction cycle [10]. In particular, the link between KCa2.3 and the heart is of interest since a genome-wide association study (GWAS) identified an association of the KCa2.3-encoding KCNN3 gene with lone atrial fibrillation in humans [11]. In addition, pharmacological blocking of atrial KCa2 is currently studied as a way to prolong selectively atrial action potential length and thereby to stop atrial fibrillation [12].

In respect to KCa3.1, single-nucleotide polymorphisms (SNP) in the KCNN4 gene have been associated with cardiac infarction in a Japanese cohort [13] and in Crohn’s disease in the Australian and New Zealand population [14, 15]. Moreover, there are several reports, which have been reviewed in depth elsewhere, on reduced, increased, or conserved functions of KCa2.3 and KCa3.1 in cardiovascular disease and other diseases such as chronic inflammation, allergy, organ fibrosis, and neurodegeneration [2–4, 16–18]. This assigns KCa2.3 and KCa3.1 as disease-relevant channels in idiopathic disease and targets in translational medicine.

The second subgroup of KCa channels, which is phylogenetically distantly related to the first subgroup [9, 19], consists of the well known KCa with large unitary conductance (200–300 pS; gene name KCNMA1, protein name: BKCa, BigK, Maxi K, or KCa1.1), and three other members (KCNU1, KCNT1, and KCNT2), of which KCNT1 and KCNT2 (protein names: SLACK and SLICK) are in fact more distantly related Na-activated K channels (KNa) with large conductance (150–300 pS) [20].

In contrast to KCa2 and KCa3.1 channels, KCa1.1 contains a voltage-sensor in S4 producing gating by depolarization, particularly in a membrane potential range, in which voltage-dependent L-type Ca2+ channels operate (Fig. 5.2b). The KCa1.1 channel complex recruits four beta-1 subunits that sensitize channel gating to Ca2+ and voltage by favoring channel opening at physiological internal Ca2+ levels reached in the vicinity of the KCa1.1 channel Ca2+ sensors.

In respect to gene expression of channels of this subfamily in the vasculature, KCa1.1 is predominant in smooth muscle and the CNS while other members (SLICK and SLACK) are predominantly expressed in neuronal tissues. KCa1.1 forms complex functional units (“signalosomes”) with physical interactions with voltage-regulated L-type Ca2+ channels [21, 22]. The electrical membrane events of such co-operations are spontaneous-transient-outward-currents (STOCs, mediated by KCa1.1) that are triggered by spatially co-localized Ca2+ influx through L-type Ca2+ channels and small Ca2+ release events from the sarcoplasmic reticulum, termed “sparks” [22, 23]. Physiologically, these KCa1.1-mediated STOCs are considered to provide a negative feedback on Ca2+ influx through L-type Ca2+ channels and thereby to limit depolarization and contractions or vasospasm [21, 24]. Presumably, diffusible EDHFs, like epoxyeicosanoids augment KCa1.1-mediated STOCs and by this produce their vasorelaxing actions.

In contrast to KCa3.1 and KCa2.3 channels, KCa1.1 does not seem to be expressed under physiological conditions in human, rat, and murine arterial endothelium, although porcine endothelium could be an exception [25, 26].

From the epidemiological perspective, SNPs in the KCa1.1-encoding KCNMA1 gene causing “loss-of-function” or in the associated beta-subunit gene (KCNMB1) causing “gain-of-function” are associated with systolic hypertension (KCNMA1) or resistance to hypertension, respectively [14, 27, 28].

KCa and Their Functions in Arterial Endothelium

In early nineties, the genes encoding for channels were still awaiting cloning (KCa3.1 in 1997 [29] and KCa2 in 1996, [30]) and KCa channels were pharmacologically characterized in endothelium and differentiated according to their sensitivity to the peptide toxins, apamin (KCa2), charybdotoxin (KCa3.1 and KCa1.1), and iberiotoxin (KCa1.1), and the small molecules, clotrimazole (KCa3.1) [31], and paxilline (KCa1.1) [32]. KCa2-blocking compounds like UCL-1684 [33] have not been invented yet. Notably, in 1996, Marchenko and Sage were first on providing comprehensive data on functional KCa3.1 and KCa2 by elegant in-situ patch-clamp electrophysiology on intact endothelium of isolated rat aorta [34]. Almost contemporaneously, KCa3.1 currents were identified in endothelia from different species and different vascular beds [2, 3, 35], for instance, in native bovine and rat aortic endothelial cells [36, 37], and rat brain capillaries [35]. In general, expression of KCa3.1 and KCa2.3 is a common finding for endothelia of several species and is documented throughout the entire vascular tree, with no considerable differences (at least qualitatively) between arterioles, arterioles of the microcirculation, and veins.

Moreover, alterations of KCa3.1 currents have been documented in a model of experimental hypertension, the spontaneously hypertensive rat [37]. This provided early electrophysiological evidence that endothelial KCa3.1 channels were altered in a hypertension-associated fashion. Moreover, these findings point to a capacity of endothelial channels to respond in a probably compensatory way to hypertension in this model. A few years later, the genes encoding for KCa3.1 and KCa2.3 currents were identified by others [26, 38] and our group as KCNN4 and KCNN3 using a single-cell RT-PCR approach combined with patch-clamp in endothelium from rats [39] and—importantly from the translational perspective—in human mesenteric endothelium [40, 41].

KCa and Endothelium-Dependent Vasodilation

In the eighties, combinations of peptide toxins in cardiovascular pharmacology were widely used to study endothelium-dependent vasodilator mechanisms by classical in-vitro myography on large rodent arteries. A contribution of KCa3.1 and KCa2 to the “EDHF” has been deduced from the blocking actions of the combination of charydotoxin (KCa3.1 and KCa1.1) and apamin (KCa2). A contribution of KCa1.1 has been discarded as iberiotoxin (KCa1.1) cannot replace charybdotoxin [4]. Yet, these peptide blockers are not fully selective and e.g. charybdotoxin—besides KCa3.1 and KCa1.1—blocks some Kv channels that are present in arterial tissue. Interestingly, in coronary arteries of pigs, “EDHF” required reportedly smooth muscle KCa1.1 as the response was abrogated by iberiotoxin, which suggested KCa1.1 as target of a diffusible EDHF (Fig. 5.1) that has been later identified as a P450-derived eicosanoids [42]. Still, vasorelaxation that involves endothelial KCa3.1 and KCa2.3 is also present in porcine coronary artery as concluded from more recent studies by others and us [43, 44].

It is noteworthy that P450-products and ω3 and ω6 fatty acids that activate KCa1.1 and have been used as antagonist of EDHF-signaling—like 20-HETE [45]—inhibit KCa3.1 [46], suggesting interactions between putative EDHF and EDH mechanisms (Fig. 5.1). Despite this evidence from in-vitro experimentation, the true contribution of EDH or possibly EDHF to circulatory regulation in humans is not established at present and a matter of ongoing debate. A main reason for the lack of knowledge is that we miss sufficiently selective, potent, safe, and ethically approved pharmacological tools for in-vivo investigation. Yet, in the following paragraphs we explore potential utilities of novel small molecule activators and inhibitors.

Endothelial Dysfunction Caused by Genetic KCa3.1/KCa2.3-Deficiency

Since 2003, the availability of gene-targeted mice as tools provided more definite experimental evidence for the significance of KCa3.1 and KCa2.3 channels for EDH-type dilation in-vitro and in-vivo [47–49]. Indeed, myography on conduit arteries and intra-vital microscopy on resistance-size skeletal muscle arterioles of KCa3.1−/−/KCa2.3T/T + Dox mice with genetically encoded deficiency of both endothelial channels revealed that loss of KCa3.1 and KCa2.3 impairs strongly endothelial and smooth muscle membrane hyperpolarization and EDH-type dilations in both vessel types [50–52] (Fig. 5.3a). Interestingly, also NO-type dilations were disturbed in the mice, suggesting that KCa3.1 and KCa2.3-mediated hyperpolarization promotes Ca2+ influx and thereby stimulates also Ca2+-dependent NO production [50, 51, 53]. A candidate channel mediating the Ca2+ influx during hyperpolarization is the TRPV4 channel [54–56], which conducts considerably Ca2+ at clearly negative potentials while it shows negative-feedback regulation by Ca2+ at potentials near zero. In support of the above-mentioned, loss of KCa3.1 in KCa3.1−/− and pharmacological inhibition of TRPV4 in mesenteric artery impaired Ca2+ dynamics to acetylcholine [55].

Fig. 5.3
figure 3

(a) On left: Schematic illustration of the impaired EDH-type arterial dilation to acetylcholine in KCa3.1 and/or KCa2.3-deficient mice. On right: Impaired shear stress-induced EDH-type dilation in KCa2.3-deficient mice. (b) Major cardiovascular, anatomical, and behavioral phenotypes. The asterisks (*) indicate phenotypes that are unaltered by subchronic treatment with doxycycline to suppress KCa2.3-gene expression. Note that KCa3.1−/− exhibit mild systolic hypertension only during locomotor activity

Full size image

Moreover, our studies on single and double KCa3.1/KCa2.3-deficient mice [50] revealed subtype specific roles of the channels for EDH-type dilation since KCa3.1 is of greater importance than KCa2 for acetylcholine-induced dilations while KCa2.3 is more important in dilations caused by mechanical stress acting on the endothelium (shear stress), for hyperemia during skeletal muscle twitching [52] and for a basal tonic dilator input from the endothelium [49]. Thus, a KCa2.3-EDH system is likely needed to regulate basal blood flow and the higher blood flow demand during exercise (metabolic demand) while KCa3.1-EDH is linked to muscarinic receptor stimulation and ER Ca2+ release and to possibly sympathetic drive on resistance arteries.

Some of these subtype specific roles can be explained by a differential intracellular localization of KCa3.1 and KCa2.3 in the endothelial cell (Fig. 5.1). However, there are some caveats because of the uncertain specificity of antibodies used to localize KCa3.1 and KCa2.3 and of lack of approach validation in KO-controls. Regardless of these uncertainties, KCa3.1 has been localized in proximity to protrusions of the endoplasmic reticulum (ER) [57] where it senses Ca2+ release events after e.g. muscarinic receptor (M3) stimulation by acetylcholine. Moreover, KCa3.1 has been reported to co-localize with myo-endothelial gap-junction proteins thus forming a functional unit for electro-tonic spread of EDH to smooth muscle (for review [2, 58–60]).

KCa2.3 has been found at intercellular junctions [61] where it is suggested to form a functional unit with TRPV4 [62] or possibly with another Ca2+-permeable channels. Interestingly, such a functional unit has recently been suggested to occur in the tubular system of the kidney [63] as well as in pulmonary and mesenteric arteries [62, 64], where TRPV4 uses KCa2.3 to produce vasodilation. Nevertheless, we need more definite visualization of KCa3.1 and KCa2.3 using gold-labeled AB, KO-controls, and transmission electron microscopy to ascertain the present view.

Interplay of KCa3.1/KCa2.3 Channels with Ca2+-Permeable Channels

Evidence from patch-clamp studies point at close functional interactions of endothelial KCa with Ca2+-permeable channels (Fig. 5.1): While patch-clamping the endothelium of intact vessel preparations or bovine endothelial cells in a flow chamber, we frequently observed that KCa3.1 gating and membrane hyperpolarization occurs following stimulation of shear stress-activated or hyposmotic stress-activated Ca2+ permeable channels [65]. Moreover, these studies showed temporally fluctuating Ca2+ release events that amplified KCa3.1 activation. These fluctuating Ca2+ release events and concomitant KCa3.1-mediated currents and hyperpolarization were modulated in frequency and amplitude by the degree of shear force acting on the cells. Interestingly, these fluctuating currents shared some similarities with spontaneous transient outward currents (STOC) seen in smooth muscle, although the endothelial STOCS were of longer duration and largely voltage-independent and did not involve KCa1.1 and KCa1.1-interacting L-type Ca2+ channels as in smooth muscle. Today, one of the candidate MSCs triggering Ca2+ signaling events and co-activation of KCa3.1 (and KCa2.3) to flow or shear stress stimulation is (again) TRPV4 as concluded from elegant in-situ imaging in cerebral and mesenteric arteries [66] and by the loss of shear stress-induced dilation in carotid arteries from TRPV4−/− mice [67].

Thus, from the perspective of integrative cardiovascular physiology, the interplay of TRPV4 and KCa2.3 (KCa3.1) can be mechanistically important to fine-tune EDH and arterial tone in response to alterations of blood flow in e.g. the working skeletal muscle or the brain with their metabolic demands [52]. In terms of cross-reactivity with other vasodilator systems, TRPV4 has been proposed to physically interact with the scaffold protein caveolin-1 and thus the endothelial nitric oxide synthase (eNOS) [68] and the Ca2+ influx through TRPV4 further stimulates Ca2+-dependent NO synthesis [54] in addition to EDH, producing some redundancy in endothelial vasoregulator capacity and during exercise.

It should be noted that KCa3.1 and KCa2 channels are not considered “smooth muscle” channels. In fact, according to our electrophysiological studies, KCa3.1 channels are absent in contractile i.e., differentiated, vascular smooth muscle of rodents and this makes some physiological sense because the channels are highly Ca2+-sensitive because of the Ca2+ sensor, calmodulin, and, moreover, are voltage-independent. Functioning of KCa3.1 in smooth muscle would almost immediately abrogate any contraction because the relatively high Ca2+ signaling in contracting smooth muscle would cause strong KCa3.1 activation and lasting hyperpolarization, which in turn causes cessation of Ca2+ influx. In other words, KCa3.1 would provide a too strong negative feedback on Ca2+-dependent contraction. In contrast, the voltage-dependent and less Ca2+-sensitive KCa1.1—known as smooth muscle channel—is the better channel to mediate negative feedback during depolarization. With respect to KCa2 it is—because of the same theoretical considerations—rather unlikely that these closely related and similarly voltage-independent KCa2 channels are functionally expressed in contractile smooth muscle. We have not been successful in detecting KCa2 channel functions in vascular smooth muscle of the rat and mouse.

Systemic Cardiovascular Alterations in KCa3.1/KCa2.3-Deficient Mice

That KCa3.1 and KCa2.3 channels add significantly to systemic cardiovascular regulation has been suggested by the higher blood pressure in KCa3.1−/− mice [50] and KCa2.3-deficient mice (KCa2.3T/T + Dox) [49, 50] as found by blood pressure telemetry in the freely moving animals (Fig. 5.3b). The hypertension has been reported to be, however, mild and overt only during physical activity (for review see [7]). Still, the latter observation was particularly interesting as it could mean that KCa3.1 and KCa2.3 and the EDH-system are needed to adjust tone during exercise and sympathetic input while they are of less or no importance during rest. Recently, KCa3.1−/− and connexin(Cx)37−/− mice—Cx37 that has been considered a major constituent of myo-endothelial gap junctions—have been shown to exhibit lower ADP/ATP-receptor (P2Y2/4)-mediated fast depressor responses and secondary increases in pressure [69] in anesthetized mice, suggesting decreased responsiveness to evoked endothelial function. Responses to acetylcholine-infusion has been reported to also be reduced at low but not high concentrations, suggesting endothelial dysfunction in vivo as it has been observed so far only in isolated KCa3.1−/− arteries.

Regarding the contribution of nitric oxide in these mice, inhibition by L-NAME has been shown to be still capable to increase blood pressure in KCa3.1−/−/KCa2.3T/T + Dox mice [7], indicating that NO signaling is largely intact in these mice and it is therefore likely that NO- and EDH mechanisms act independently from each other to a large extent but are likely additive.

Other Phenotypes

It is worth to mention that KCa3.1−/− mice have been shown to exhibit spontaneous physical hyperactivity (but without signs of general distress) and exhibited complex alteration in monoamine levels, with increased noradrenalin turnover in frontal cortex and lower serotonin-turnover in frontal cortex, striatum and brain stem, and no change in dopamine turnover [70]. Plasma corticosterone has been found to be normal but increased to higher levels in KCa3.1−/− under stress as reported by others [71]. This suggests substantial alteration of neurotransmission and behavior in KCa3.1, which can add to the cardiovascular phenotypes in these mice. Another solid phenotype in KCa3.1−/− is progressive splenomegaly (Fig. 5.3b) caused presumably by defects in erythrocyte volume regulation and a resulting increased erythrocyte degradation and turnover [72], producing a higher workload for the spleen and secondary adaptation of size to meet the demand.

Also KCa2.3T/T (+Dox/−Dox) have complex behavioral and morphological phenotypes (Fig. 5.3b) [1, 70], which can be linked in KCa2.3T/T−Dox to larger afterhyperpolarization, refractory periods, and thus firing frequencies. Locomotor hypoactivity is evident in KCa2.3T/T+Dox and −Dox (Fig. 5.3b). Moreover, there are significant anatomical (developmental) changes in the vasculature and the heart, i.e. visible enlargement of diameters of mesenteric and uterine arteries and an enlarged heart with increased right ventricular wall thickness and altered atrioventricular morphology [49, 73, 74]. Postnatal development of retinal vessels has been found to be normal, although deficiency of KCa2.3 (KCa2.3T/T+Dox) increased branching (D. Rappert, 2011, German MD thesis; archiv.ub.uni-marburg.de/diss/z2011/0306/pdf/ddr.pdf). In sum, these anatomical changes suggest that alterations of KCa2.3 expression levels can have a substantial impact on the development of the vasculature. As possible pathophysiological consequences, the mice experience parturition defects [48] and sudden cardiac death [75] when over-expressing the channels (KCa2.3T/T−Dox).

KCa3.1 and KCa2.3 in Cardiovascular and Other Disease States

There is growing evidence that KCa3.1 and KCa2.3 channels are differentially regulated in disease. In fact, up-regulated KCa3.1 functions have been found to promote many disease states characterized by abnormal proliferative organ remodeling and inflammation such as neointima formation [76, 77], atherosclerosis[78], fibrosis of lungs [79], fibrosis in damaged and diabetic kidneys [80, 81], possibly liver cirrhosis and portal hypertension [82], allograft vasculopathy [83], kidney allograft rejection [84], as well as in obliterated trachea transplants [85]. Such up-regulation of KCa3.1 in diseased tissues reflected infiltration of KCa3.1-expressing T cells and macrophages and on de-novo mRNA-expression of KCa3.1 in the tissue itself. The de-novo mRNA-expression was induced by actions of classical growth factors on smooth muscle, fibroblast, and endothelial cells in vitro [80, 86, 87]. Moreover, shear stress has been shown to induce KCa3.1 in human endothelial cells [88]. In general, the induction of KCa3.1-mRNA synthesis requires activation of the MEK/ERK MAP kinase cascade and transcriptional mechanisms involving activator protein-1 (c-jun) and repressor protein-1 as negative regulator of gene expression [87]. Also the Ca2+/calmodulin-dependent protein kinase kinase/Akt/p300 cascade has recently been reported to mediate the up-regulation of KCa3.1 in endothelial cells during shear stress stimulation [89].

With respect to endothelial functions and cardiovascular disease, endothelial KCa2.3 function and EDH were reduced in regenerated endothelium [90], renal insufficiency [91], ageing [92], and ovariectomy [93], which could explain some aspects of the endothelial dysfunction reported in these conditions. However, up-regulation of gene expression of the channel has been documented in obese rats [94, 95], in pulmonary hypertensive mice [74], and by shear stress [89] in vitro, suggesting here compensatory roles of the channel to maintain/improve endothelial vasodilator function under these conditions. The precise mechanisms of the altered gene expression in endothelium are not understood. However, estrogen is considered a major regulator of KCa2.3 expression and this positive regulation can be an additional explanation for pre-menopausal cardiovascular protection in women.

Interestingly, low KCa2.3 mRNA expression in mammary arteries correlated negatively (r > −0.5) with blood pressure in a small Danish cohort of patients with or without renal failure (uremia) or of diabetic and non-diabetic patients (all, n = 55) (unpublished data provided by Dr. Lars Melholt Rasmussen, Director of the Odense Artery Biobank SDU, Denmark).

In contrast, KCa3.1 expression did not correlate with blood pressure.

Regarding other tissue and disease states, it is worth to mention that, both, KCa2.3 and KCa3.1 are highly expressed in some cancers [96–99] and have been implicated in cancer progression and metastasis (for review of the field see also [100, 101]). In addition, KCa3.1 has been found to be up-regulated during immune responses and in chronic inflammation such as in T cells within the chronically inflamed mucosa in ulcerative colitis [102] and in microglia after ischemic stroke [103], which is in line with the notion that KCa3.1 is an important regulator of immune cell function [18].

Endothelial KCa3.1 and KCa2.3, Their Pharmacological Manipulation, and Systemic Cardiovascular Consequences

From the pharmacological viewpoint, the introduction of TRAM-34 as a potent and selective KCa3.1 blocker in early 2000 by the laboratory of Heike Wulff provided a “modernized” pharmacological tool to define mechanisms of EDH-type or EDHF-type dilation, in addition to TRAM-34’s value for testing roles of KCa3.1 in inflammatory and autoimmune disease [104]. TRAM-34 derived from the fungicide and P450-blocker, BayerAG compound clotrimazole that was the first well studied small molecule blocker of KCa3.1 (Fig. 5.4a). TRAM-34 is more selective for KCa3.1 since the imidazole ring in clotrimazole—being required for P450-blockade—was replaced by a pyrazole ring. TRAM-34 was initially developed as possible immunosuppressant as it blocks KCa3.1 in T cells and macrophages and thereby proliferation, migration, and cytokine production of lymphocytes [104].

Fig. 5.4
figure 4

(a) Chemical structures of selected KCa2/3.1 modulators with nanomolar potencies. (b) Upper panel: Schematic illustration of inhibition of mixed voltage-independent Ca2+-activated KCa2.3/KCa3.1 currents by TRAM-34 (KCa3.1) and UCL-1684 (KCa2.3). Lower panel: SKA-121-evoked potentiation of mixed KCa2.3/KCa3.1 currents and antagonism by the pan-negative gating modulator, RA-2, at equimolar concentration. (c) Hypothetical interaction sides of blockers and gating modulators

Full size image

The introduction of TRAM-34 into the cardiovascular field greatly helped to define with more precision the requirement of KCa3.1 for EDH-type dilation in several vascular beds and from different species, including human arteries. While it is clear that TRAM-34 reduces acetylcholine-induced vasodilations, this does not have any systemic consequences in vivo since TRAM-34 injections into mice or rats did not cause a change in blood pressure. Interestingly, blocking KCa3.1 in the blood-brain-barrier by TRAM-34 has been reported to reduce water movements into the brain and early edema formation after ischemic stroke [105]. Another structurally different KCa3.1-blocker developed by the BayerAG was effective in reducing brain edema caused by trauma [106].

Senicapoc (ICA-17043; Fig. 5.4a), another KCa3.1-blocker similar to TRAM-34 advanced into clinical trials as a treatment for sickle cell disease [107] and has been found to be cardiovascular-safe since it did not increase blood pressure in humans. Senicapoc has been proposed to be efficient in sickle cell disease because blocking of the over-active KCa3.1 may hinder irreversible erythrocyte sickling (shrinkage) by blocking volume reduction by KCa3.1-mediated K efflux. In Phase IIb and III clinical trials, Senicapo c was found to improve erythrocyte parameters and hemoglobin content, but, unfortunately, did not prevent but rather increased painful vaso-occlusive crisis as primary end point for unknown reasons. The disappointing outcome terminated the development of Senicapoc at least for this indication. Whether endothelial dysfunction was one reason why Senicapoc failed is not known and has not been studied further.

From a more general perspective, we may consider that—during KCa3.1-blockade—the other potent vasodilator system, the NO system and KCa2.3 channels are still functional and have the capability to compensate the inhibition of KCa3.1. Still, in the case of reduced NO-synthesis and/or NO-availability, inhibition of endothelial KCa3.1 (and KCa2.3) may aggravate the impaired endothelial vasodilator influence on tone and could represent a cardiovascular safety issue.

Senicapoc has been also tested as treatment for exercise and allergic asthma and failed. Still, this KCa3.1 blocker—albeit safe—may have other utilities as mild in the case of immunosuppressant in chronic inflammatory disease, e.g. chronic inflammatory bowel disease, in reducing astrocyte-mediated secondary damage after spinal cord injury [108] and microglia-mediated neurotoxicity [109] and neurodegenerative disease such as Alzheimer’s disease [110].

Recent Developments: Positive-Gating Modulators

From the perspective of cardiovascular translational pharmacology, activators of KCa2.3 and KCa3.1 may have more therapeutic utilities in cardiovascular disease characterized by endothelial dysfunction than obviously blockers. However, the existing activators with reasonable potency have not been tested in humans. At present all experimental evidence for their potential therapeutic utilities in cardiovascular disease relies on in-vitro experimentation and few in-vivo cardiovascular monitoring in rats, mice, and dogs [111–114].

The Neurosearch compound NS309 (Fig. 5.4a) has widespread use in in-vitro pharmacological experimentation. It has a high potency but a disadvantageous selectivity profile as the compound, besides activating KCa3.1 and KCa2 channels, has additional blocking effects on voltage-gated Ca2+ channels. Moreover, NS309 blocks cardiac hERG channels that may impair ventricular repolarization, which can be regarded a severe safety issue.

The development of SKA-31 (Fig. 5.4a), using the ALS-drug Riluzole (a mixed glutamate antagonist, Na+ channel blocker and unselective K+ channel activator with low potency) as template, provides first evidence that systemic activation of KCa3.1/KCa2.3 improves endothelial vasodilator function, particularly in vivo. In fact, SKA-31 displays fivefold selectivity for KCa3.1 over KCa2, channels and has been shown to potentiate endothelial KCa3.1/KCa2.3 currents as well as EDH-vasodilation to acetylcholine in carotid arteries of mice [53, 111]. In the microcirculation of skeletal muscle (cremaster muscle) SKA-31 has be shown to induce vasodilation in its own right [115], suggesting that there is some basal Ca2+ signaling or “Ca2+ pulsar” activity that gives rise to some KCa3.1 activation, which can be potentiated by SKA-31. Moreover, SKA-31 has been found to improve coronary blood flow in rats [112], to potentiate bradykinin-induced relaxation and to reduce serotonin-induced contraction in large porcine coronary artery [44]. In a model of severe fatal hypertension, SKA-31 has been found to increase renal blood flow and to increase survival in mouse model of fatal hypertension [16, 116]. In other tissues, SKA-31 modulated bladder contractility and SKA-31 decreased human detrusor muscle excitability and contractility, suggesting utility of the activator in overactive bladder [117].

Mechanistically, it should be noted, that SKA-31 and recently developed other SKAs [114] are not simple channel openers like other channel openers that activate the channel in the closed configuration (no gating). Rather, SKAs act as positive-gating modulators that keep the Ca2+-gated channel in the open-configuration and thereby shift Ca2+ dependence of the channel to the left [114] (Fig. 5.5a). In other words, the channel has a higher activity (open probability) at Ca2+ concentrations that are normally not sufficient to produce large channel activation on its own.

Fig. 5.5
figure 5

(a) Schematic illustration of increase of Ca2+ sensitivity of KCa3.1-activity (open probability) by SKAs (left shift of the concentration-response curve). RA-2 produces a right shift of the curve, indicating reduced Ca2+ sensitivity. (b) Schematic illustration of SKA-induced potentiation of bradykinin-induced dilation of porcine coronary artery and antagonism by RA-2. (c) On left: Illustration of blood pressure lowering effects of SKA-121. RA-2 has no appreciable effect. On right: Mild bradycardia induced by SKA-121 and appreciable bradycardia evoked by RA-2

Full size image

The very elegant work by Michael Zhang’s group on crystals of the KCa2.2 c-terminus has provided first insight into molecular details and potential bindings domains. They showed that 1-EBIO, an early less potent precursor of DC-EBIO, fits into a pocket between calmodulin and the c-terminus [118] (Fig. 5.4c). Moreover, they showed that NS309, the other activator of this compound class, stabilizes such interactions by stabilizing an intrinsically unordered linker between the CAM-binding domain to S6, which can explain the molecular mechanics of channel activation ([119] for details; for review see [120]).

What would be the advantage of a positive-gating modulator compared to a more classical activator such as e.g. activators of KATP channels in cardiovascular disease? A mechanistic advantage of the KCa2/3 activators is presumably that they act only when there is Ca2+ mobilization, e.g. in form of “Ca2+ pulsars” in the endothelium [57] and thus initiation of endothelial function. Here, we would stimulate the active endothelium or potentiate endothelial function “when it is needed”. Indeed, at present we understand well how the endothelium regulates arterial tone in-vitro and inhibition of NO synthesis in-vivo clearly elevates blood pressure while NO-donors are in the clinics since long to alleviate angina pectoris and hypertensive crisis. This fosters the pivotal role of NO in the systemic circulatory regulation. However, we do not know when endothelial EDH-vasodilator function occurs in the organisms and under physiological conditions and what are the consequences systemically or locally.

In situation of endothelial dysfunction/degeneration with loss of both channels an activator would not make sense at all while some conserved KCa3.1/KCa2.3 functions could be potentiated and improve thereby endothelial vasodilator function.

Systemic Effects of Positive-Gating Modulation

What do we learn from cardiovascular telemetry? In freely moving unstressed mice that have a blood pressure similar to ours (around 120/85 mmHg Systole/Diastole), acute administrations of SKA-31 (30–100 mg/kg; giving plasma levels above the EC50 for KCa3.1 and KCa2 activation) caused a rapid drop (by approx. −30 mmHg) in systolic and diastolic pressures that persisted over 1–4 h, pending on dose [111, 115]. In hypertensive mice, SKA-31 was similarly effective and lowered pressure to normotensive levels (mice treated with L-NAME or connexin(Cx)40−/− with angiotensin-II hypertension). Moreover, intra-vital microscopy on the microcirculation of cremaster skeletal muscle in anesthetized mice reveals that SKA-31 was capable to produce substantial arteriolar dilation from basal tone that did not require NO, but was abolished by genetic KCa3.1 deficiency [115]. This effect in resistance-size arteries explains to some extent the efficiency of SKA-31 to produce hypotension or normotension from hypertensive levels. Thus, this positive-gating modulator of KCa3.1 and KCa2.3 is active and causes the expected decrease in blood pressure.

A major concern is, however, that the pressure drop in the mice is accompanied by strong bradycardia (300 bpm at 100 mg/kg vs. 600 bpm (normal)) [115]. This is also seen in KCa3.1−/−, suggesting that this bradycardia is not related to activation of KCa3.1 but was rather caused by KCa2 activation in atria of the heart ensuing loss of sinus rhythm, and/or block of transmission within the conduction system. Interestingly, genetically encoded KCa2.3 over-expression has been shown to increase atrioventricular refractory period in young KCa2.3-overexpressor (KCa2.3T/T) but decreased it in old KCa2.3T/T mice because of anatomical alterations related to KCa2.3-over-expression [75]. Moreover, genetic deficiency of the KCa2.2 subtype causes prolonged the PR and RR intervals [121] and caused prolongation of atrioventricular transmission, while over-expression of KCa2.2 had opposite effects.

Together, the data makes it likely that SKA-31 at KCa2-activating plasma concentrations affects the conduction system of the murine heart, which, unfortunately, may hinder a further development of this compound. Another, drawback is that SKA-31 causes severe sedation (immobility) in a KCa3.1-independent fashion [70], which relies likely on the bradycardia and/or the activation of central KCa2 channels, slowing neurotransmission, or—possibly—skeletal muscle KCa2, producing paresis.Footnote 2

Interestingly, dogs that have a heart rate similar to us respond differently to SKA-31 (injected i.v.) because they show at KCa3.1/KCa2-activating plasma levels profound but short-lived hypotension and reflex tachycardia [113], which may point to substantial species differences and request the need for thorough cardiovascular safety monitoring in large mammals.

Yet, SKA-31 and an analogue of SKA-31, SKA-19 has been found effective in the NINDS-funded anti-convulsant screening program [123]. Nonetheless, the likely disadvantageous cardiovascular profile of SKA-31 generates the need of a more selective KCa3.1-activator. Heike Wulff’s group at UC-Davis synthesized a series of SKA-analogues [123], of which SKA-121 (5-methylnaphtho[2,1-d]oxazol-2-amine; Fig. 5.4a) has an improved selectivity profile for KCa3.1 (approx. 40-fold higher for KCa3.1 over KCa2). At the level of endothelium-dependent vasorelaxation, we find that SKA-121 does not act as vasorelaxant on its own in large coronary artery of the pig but potentiates the response elicited by bradykinin (Fig. 5.5a). This can be reversed by TRAM-34 [123]. Moreover, the compound does not affect endothelium-independent contraction or relaxation. In telemetry (Fig. 5.5b), SKA-121 reduced blood pressure in normotensive mice and L-NAME-treated hypertensive mice over approx. 2 h and 6 h, respectively, but did not cause blood pressure alterations in KCa3.1−/− mice. Importantly, heart rate was insignificantly affected in this study, suggesting substantial improvement of selectivity and cardiovascular safety of SKA-121. Neurological effects still need to be investigated in details, but sedation can still be an issue because of the higher brain/plasma concentration ratio [123]. In sum, positive gating modulators like SKA-121 are likely to have some therapeutic utilities in cardiovascular disease (hypertensive crisis, vasospasm, and central and peripheral ischemia) (Fig. 5.6).

Fig. 5.6
figure 6

Schematic illustration of systemic actions of the positive- and negative gating modulation of KCa2/KCa3.1 channels

Full size image

Recent Developments: Negative-Gating Modulators

Besides the potential utility of positive-gating modulators of KCa3.1/KCa2 in cardiovascular pathophysiology—albeit counterintuitive at first—also negative-gating modulators of KCa3.1/KCa2 channel activators are of potential pharmacological value, for instance in atrial fibrillation, hypotension and sepsis, hyperemia, atherosclerosis, and restenosis after balloon catheter intervention, but perhaps also in neurological disorders associated with microglia-activation, chronic inflammation, and possible some cancer over-expressing KCa3.1/KCa2 channels.

The present selective KCa3.1/KCa2 inhibitors are mainly blockers that obstruct ion flow at the intracellular cavity below the selectivity filter (such as TRAM-34 on KCa3.1, Fig. 5.4c) or from the outside as classical pore blockers such as UCL-1684 (Fig. 5.4a, c), structurally mimicking the peptide blocker, apamin. The utility of UCL-1684 for in-vivo experimentation is not clear, but this very large molecule is still not drug-like. TRAM-34 has been used frequently in experimental in-vivo intervention trials and has been proven to be effective to inhibit BCI-induced neointima formation in rats and pigs [76, 77], neo-angiogenesis in Matri-gels [124], and experimental fibrosis in normal and diabetic kidneys [80, 81] and lungs [79, 125], suggesting utilities of KCa3.1-inhibitors for the treatment of disease characterized by pathological cell proliferation and pathological organ remodeling (for review see [8, 126]). However, TRAM-34 has been reported to be an inducer of P450 enzymes and looses selectivity for KCa3.1 over some other K channels at μmolar concentrations. Therefore, we have recently performed a small screening campaign focusing on another compound class (phenols and polyphenols with beneficial properties as anti-oxidants) and find 13b, tri-fluoro-benzoate ester [44, 127] as reasonably potent. Moreover, this compound turned out to act as negative-gating modulator because it competes (in this regard unlike TRAM-34) with the positive-gating modulators, SKA-31 and SKA-121 (Fig. 5.4a, c). Unfortunately, 13b is a large and lipophilic molecule and bioavailability of 13b is presumably poor. To generate a new more drug-like analogue, we synthesized smaller and less lipophilic variants, of which RA-2 (1,3-phenylenebis(methylene)bis(3-fluoro-4-hydroxybenzoate) (Fig. 5.4a) has acceptable drug likelihood and conserved potency (IC50 approx. 20 nM), a good selectivity over other K channel families, but also blocks with similar potencies all three KCa2-subtypes [128]. This can be explained by the negative-gating modulation at the level of calmodulin-activation (Fig. 5.4c), which is alike in the complete KCa3.1/KCa2 family. As to be expected for a negative gating modulator, RA-2 shifts the concentration-response curve for Ca2+ activation to the right (Fig. 5.5a).

Thus, RA-2 is the first pan-negative gating modulator of KCa3.1 and KCa2. In in-vitro experimentation (Fig. 5.5b), RA-2 has been found to inhibit bradykinin-induced EDH-type relaxation in porcine coronary artery and antagonizes potentiation of the bradykinin response by the positive-gating modulator, SKA-121 [128].

Systemic Effects of Negative-Gating Modulation

In respect to systemic cardiovascular regulation, RA-2 appears to be relatively safe since acute i.p. injections of up to 100 mg/kg did not produce hypertension or any disability in the mice (Fig. 5.5b). Still, RA-2 caused lasting bradycardia (drop by 150 bpm from the high levels of approx. 600 bpm, Fig. 5.5c). The simplest explanation for this bradycardia is that it mirrors baroreceptor-reflex to a higher peripheral resistance caused by KCa3.1 and KCa2.3 inhibition in small resistance-size arteries (Fig. 5.6). Other possibilities are a direct effect on the heart conduction system, caused by action potential prolongation as a result of KCa2 inhibition. However, we do not wish to exclude other changes in sympathetic or parasympathetic drive on the heart or central and peripheral mechanisms. Interestingly, bradycardia was absent in KCa3.1−/− suggesting an effect that is selectively mediated by KCa3.1. Besides these cardiovascular actions of RA-2, our unpublished data demonstrate a higher locomotor activity in RA-2 treated mice, which is, interestingly, also a feature of KCa3.1−/− mice and fosters the view of a participation of KCa3.1 in control of behavior and/or physical activity (Fig. 5.6).

Concluding Remarks

The current evidence derived from experimentation on gene-targeted mice and advance KCa3.1/KCa2.3 pharmacology assigns substantial impact of endothelial KCa3.1 and KCa2.3 in the endothelium on local arterial regulation as well as systemic cardiovascular regulation. Here, EDH and to some extent also NO are likely the downstream effectors. However, with respect to humans, we still need to be careful since the contribution of the channels to human systemic cardiovascular regulation remains unexplored although there is no doubt that human endothelium expresses these channels.

KCa3.1 and KCa2.3 are differentially regulated by cardiovascular disease at the functional and gene expression level and by other disease states. However, “loss-of-function” or “gain-of-function” (in the sense of monogenetic channelopathies) are unknown so far. Still, the KCa2.3-gene is linked to atrial fibrillation and KCa2 blockers are currently developed by the Danish Spin-off Acesion as novel types of atria-selective antiarrhythmic drugs.

Idiopathic forms of cardiovascular dysfunction in, particularly, human subjects go along with altered EDH and variable changes of KCa3.1/KCa2.3 gene expression and functions that could point at compensatory or pathogenic roles. This offers venues for endothelial selective treatment of endothelial dysfunction by activators like the positive-gating modulator, SKA-121 (Fig. 5.6). While the development of KCa3.1/KCa2 channel activators as antihypertensive is unlikely considering the availability of several established medications including Ca2+ antagonists and the—for activators—potentially problematic tachyphylaxia, they may still be of use as alternative for the treatment of angina pectoris, local peripheral ischemia and pain, intra-surgical hypertension, and vascular protection of transplants. Not to forget that KCa3.1/KCa2-activators may be of advantage in early ischemic stroke and cause neuroprotection by impeding Ca2+/Na+ overload in hypoxic neurons. Besides, such utilities in cardiovascular and cerebrovascular disease, they may also serve as targets in neurons to treat epilepsy and pain (Fig. 5.6).

In contrast, KCa3.1 and KCa2 inhibitors such as the pan-negative gating modulator of this channel class, RA-2, could have therapeutic value in situation of systemic hypotension as it occurs during sepsis, anesthesia accident, or after resurrection (Fig. 5.6). Moreover, it may have utilities for chronic inflammatory processes and proliferative angiopathies and, if considering KCa2/3.1 expression in brain and skeletal muscle, for the treatment of motivation loss and muscle weakness (Fig. 5.6).

Yet, these gating-modulators of KCa3.1 and KCa2 channels may be of help to shed new light on physiological and pathophysiological roles of KCa3.1 and KCa2 channels in the organisms. Considering the complex tissue expression pattern of the channels, the efficacy or undesired side effects of pharmacological manipulation remains to be tested in more detail. Still, the broad therapeutic utilities of KCa3.1/KCa2 gating modulation offers several attractive venues for further pharmaceutical development.