Endothelial Small- and Intermediate-Conductance KCa Channels: An Update on Their Pharmacology and Usefulness as Cardiovascular Targets

Abstract

Most cardiovascular researchers are familiar with intermediate-conductance KCa3.1 and small-conductance KCa2.3 channels because of their contribution to endothelium-derived hyperpolarization (EDH). However, to immunologists and neuroscientists these channels are primarily known for their role in lymphocyte activation and neuronal excitability. KCa3.1 is involved in the proliferation and migration of T cells, B cell, mast cells, macrophages, fibroblasts and dedifferentiated vascular smooth muscle cells and is, therefore, being pursued as a potential target for use in asthma, immunosuppression, and fibroproliferative disorders. In contrast, the three KCa2 channels (KCa2.1, KCa2.2 and KCa2.3) contribute to the neuronal medium afterhyperpolarization and, depending on the type of neuron, are involved in determining firing rates and frequencies or in regulating bursting. KCa2 activators are accordingly being studied as potential therapeutics for ataxia and epilepsy while KCa2 channel inhibitors like apamin have long been known to improve learning and memory in rodents. Given this background, we review the recent discoveries of novel KCa3.1 and KCa2.3 modulators and critically assess the potential of KCa activators for the treatment of diabetes and cardiovascular diseases by improving endothelium-derived hyperpolarizations.

INTRODUCTION

The human genome contains eight Ca²⁺-activated K⁺ channels, which can be divided into two well defined but only distantly related groups based on their genetic relationship, single channel conductances and molecular mechanisms of calcium-“sensing”¹. The first group consists of the large-conductance Ca²⁺-activated KCa1.1 channel, also known as BK or Maxi-K, and the related KCa4.1 (Slack), KCa4.2 (Slick) and KCa5.1 (Slo3) channels, which are activated by increases in intracellular Na⁺ and/or Cl⁻ or by alkalization. The second group of channels, the KCa2/3 group, comprises the three small-conductance KCa2 channels, KCa2.1 (SK1), KCa2.2 (SK2) and KCa2.3 (SK3), as well as the intermediate-conductance KCa3.1 (IK1) channel, which all four share the same calmodulin-mediated gating mechanism. In contrast to KCa1.1, which directly binds Ca²⁺, KCa2 and KCa3.1 channels are constitutively associated with calmodulin^2,3, which functions as their Ca²⁺-sensing β-subunit, and induces Ca²⁺-dependent channel opening with reported EC₅₀ values ranging from 95 to 350 nM for KCa3.1 and from 320 to 750 for the KCa2 channels¹. Another distinguishing biophysical characteristic of the KCa2/3 channels is that their S4 segment contains fewer charged residues than the S4 segment of KCa1.1 or Kv channels. This results in a lack of voltage-dependence, which enables these channels to remain open at negative membrane potentials and to thus hyperpolarize the membrane towards values near the K⁺ equilibrium potential of −89 mV. KCa3.1 and KCa2 channels are accordingly expressed in cells that need to be able to hyperpolarize in order to 1) sustain Ca²⁺ influx through inward rectifier Ca²⁺ channels like CRAC (Ca²⁺ release activated Ca²⁺ channel) during cellular activation and proliferation; 2) to regulate firing frequency by preventing an untimely or premature action potential initiation; or 3) to pass on hyperpolarization through gap junctions.

Modulation of Ca²⁺ influx

The first function, modulation of Ca²⁺ influx during cellular activation and proliferation together with a role in erythrocyte volume regulation and fluid and electrolyte secretion, which will not be discussed in this review and which the interested reader can find reviewed elsewhere^4,5, is primarily fulfilled by KCa3.1. The channel is accordingly widely expressed throughout the body and found in cells of the hematopoietic system (i.e erythrocytes, platelets, T cells, B cells, mast cells, monocytes/macrophages, microglia); epithelial tissues in the lung and gastrointestinal tracts, as well as in vascular endothelial cells, fibroblasts and proliferating neointimal vascular smooth muscle cells^5–10. In these cells KCa3.1 regulates Ca²⁺ entry and thereby modulates Ca²⁺ signaling as depicted in the cartoon in Figure 1. Following increases in intracellular Ca²⁺ after, for example T cell receptor engagement in T cells or FGF receptor activation in fibroblasts, KCa3.1 generates a negative membrane potential, which sustains calcium entry through CRAC or transient receptor potential (TRP) channels^11,12. KCa3.1 thus drives cellular proliferation in T and B cells, fibroblasts and neointimal vascular smooth muscle cells, while KCa3.1 blockers accordingly inhibit the proliferation of these cell types^7–10.

FIGURE 1.

Cartoon of the physiological role of KCa3.1. The channel is activated by increases in intracellular Ca²⁺ following Ca²⁺ release from the ER (endoplasmatic reticulum), and/or Ca²⁺ influx through inward-rectifier Ca²⁺ channels like CRAC (Ca²⁺ release activated Ca²⁺ channel) or TRP (transient receptor potential) channels. PLC; phospholipase C; IP₃ inositol-triphosphate; CAM; calmodulin.

Regulation of firing frequency

In contrast to KCa3.1, which following its cloning in 1997^13–15 was initially assumed to be completely absent from neuronal tissue but has recently been described to be expressed in cerebellar Purkinje cells where it apparently is involved in suppressing low frequency parallel fiber input¹⁶, KCa2 channels are widely expressed throughout the CNS and are probably best known for their contribution to the medium afterhyperpolarization (mAHP) in neurons^17,18. While KCa2.1 and KCa2.2 are particularly enriched in the cortex and the hippocampus, KCa2.3 is most prominent in subcortical areas like the striatum, thalamus and monaminergic nuclei^17,19,20. By contributing to the mAHP, KCa2 channels regulate neuronal firing frequency for example in hippocampal CA1 pyramidal neurons^17,21–23, in the subthalamic nucleus²⁴, as well as in Purkinje neurons²⁵. Further deserving mention is the overlapping expression of KCa2.3 and dopamine synthesizing enzymes²⁶. The autonomous pacemaker and bursting activity of dopaminergic neurons in the substantia nigra has accordingly been shown to be largely regulated by KCa2.3^27,28. In addition to these roles in regulating somatic excitability, KCa2 channels have also been found to be localized in the postsynaptic membrane of glutamatergic synapses, where their activation and regulated trafficking modulates synaptic plasticity, thereby affecting learning and memory (for a recent excellent review please see Adelman 2012¹⁸).

KCa2 channel activity intimately follows the increases and decreases in free Ca²⁺ concentration originating from nearby Ca²⁺ sources (Figure 2). Under “normal” conditions, when neuronal [Ca²⁺]_i is not pathologically increased, the Ca²⁺ for the activation of KCa2 channels has been shown to come from several different sources: 1) voltage-gated Ca²⁺ channels; 2) Ca²⁺ permeable ligand-gated ion channels such as NMDA or nicotinic acetylcholine receptors (nAChR); 3) IP₃ mediated Ca²⁺ release from the ER downstream of G-protein coupled receptors; or 4) Ca²⁺-induced Ca²⁺ release from ryanodine receptors¹⁸. For example, in CA1 hippocampal neurons KCa2 channels and L-type Cav channels have long been known to be located in microdomains in which they are only 50–100 nm apart²⁹, while in cerebral Purkinje neurons KCa2 channels are coupled to P/Q type channels³⁰. In contrast, in dendritic spines of hippocampal or amygdala neurons, KCa2 channels seem to form feedback loops with NMDA receptors and are located in close proximity^31,32.

FIGURE 2.

Cartoon depicting the different Ca²⁺ sources involved in KCa2 channel activation. In neurons KCa2 channels are often localized in close proximity to NMDA receptors or to voltage-gated Ca²⁺ channels (Cav), which can induce Ca²⁺-induced Ca²⁺ release from ryanodine receptors (RyR). KCa2 channels can further be activated through G-protein induced Ca²⁺-release from inositol-triphosphate (IP₃) receptors and/or Ca²⁺ influx through TRP (transient receptor potential) channels.

Outside of the nervous system, KCa2.3 channels are also expressed in the liver, where they are believed to be involved in metabolic stress responses³³, and urinary bladder smooth muscle, where they play an important role in determining excitability and contractility³⁴. Since KCa2 channel inhibition with apamin has been reported to prolong action potential duration in both human and mouse atrial myocytes³⁵ and KCa2.3 variants have been associated with lone atrial fibrillation³⁶, KCa2 channels are further increasingly being recognized as contributors to cardiac excitability and are currently being investigated as novel targets for atrial fibrillation³⁷.

Endothelium-myocyte coupling

Although KCa3.1 and KCa2.3 typically have a very different tissue distribution (see above), one place, where they are expressed together, is vascular endothelium³⁸. This co-expression was initially simply viewed as redundancy but it is becoming increasingly clear that the two endothelial KCa channels are located in spatially distinct microdomains and apparently participate in different signaling pathways^38,39. While KCa3.1 channels are predominantly localized to endothelial cell projections traversing the internal elastic lamina and are activated by Ca²⁺ released from the ER in response to stimulation of muscarinic acetylcholine receptors and perhaps other GPCRs^40,41, KCa2.3 is found at inter-endothelial junctions and has been shown co-localize in caveolae with TRP channels⁴² and, presumably, also inward rectifier K⁺ channels⁴³. In this location KCa2.3 has been suggested to sense local Ca²⁺-increases produced by mechanical deformation during shear stress stimulation⁴⁴ and to amplify endothelial hyperpolarization by relieving endogenous blockade of inward rectifiers⁴⁵. For a more detailed review and some excellent cartoons depicting KCa3.1 and KCa2.3 localization and signaling the interested reader is referred to a number of recent reviews on EDH^38,39,46,47. However, even though KCa3.1 and KCa2.3 seem to be part of different signal transduction pathways responding to different upstream stimuli, the activation of either channel initiates hyperpolarization and subsequent endothelium derived hyperpolarization (EDH) mediated vasodilatory responses in which the endothelial hyperpolarization spreads to the underlying vascular smooth muscle cell layer, closes voltage-gated calcium channels, and finally produces relaxation and vasodilation^{38,39,46–48}. The relevance of this KCa3.1/KCa2.3 EDH system for systemic cardiovascular regulation is demonstrated by the observation of a higher blood pressure in KCa3.1 and/or KCa2.3 deficient mice⁴⁴. Even though the deficiency of either channel increases mean arterial blood pressure (MAP) by 7–9 mmHg, absence of KCa3.1 appears to increase MAP because of a higher systolic pressure (SP) and higher pulse pressure (PP) at a normal diastolic pressure (DP), while KCa2.3-deficiency increases MAP by increasing both SP and DP with no change in PP (Table 1, for extensive review see⁴⁹). Combined deficiency increased systolic and diastolic pressure as well as PP, but, surprisingly, did not increase MAP further suggesting compensation by other mechanisms (Table 1). Interestingly, systemic challenge of the mice with the NO-synthase inhibitor L-NAME, increased MAP in all mice by 6–11 mmHg and lowered heart rate suggesting a conserved ability to produce NO and intact baroreceptor functions (Table 1). An unexpected finding in these animals was that the higher MAP in either KO-mice was more pronounced during activity at night⁴⁴, suggesting a role of the endothelial KCa channels and the EDH system for adjusting blood pressure during exercise.

Table 1.

Aggravation of hypertension in KCa3.1/KCa2.3-deficient mice by Inhibition of NO-synthesis.

genotype	n	SP before	SP w L-NAME	DP before	DP w L- NAME	MAP before	MAP w L-NAME	HR before	HR w L-NAME
KCa3.1+/+/KCa2.3+/+(wt)	6	111 ± 2	124 ± 2#	83 ± 2	90 ± 2#	97 ± 1	107 ± 2#	607 ± 16	556 ± 10#
KCa3.1+/+/ KCa2.3T/Tdox	6	125 ± 3*	138 ± 4#	94 ± 1*	102 ± 3#*	110 ± 2*	119 ± 3#*	583 ± 20	528 ± 19#
KCa3.1−/−/KCa2.3+/+	4	124 ± 4*	134 ± 2#	91 ± 3	95 ± 3#	108 ± 3*	114 ± 2#	625 ± 20	543 ± 37#
KCa3.1−/−/ KCa2.3T/Tdox	6	122 ± 3*	135 ± 4#	91 ± 2*	99 ± 4#	107 ± 2*	118 ± 4#	565 ± 13	528 ± 18

SP = systolic blood pressure; DP = diastolic blood pressure; MAP = mean arterial pressure; HR = heart rate; PP = pulse pressure; Dox = doxycycline to turn off KCa2.3 expression; Telemetric data are 24-hrs means ± SEM.

^*P<0.05 vs. wild-type, One-way ANOVA and Bonferroni post-hoc test.

^#P<0.05,

24 hrs L-NAME (50 µg/ml drinking water) vs. before treatment, paired Student´s t-test.

PHARMACOLOGY

In contrast to many other ion channels, KCa3.1 and KCa2 channels have a relatively well-developed pharmacology and pharmacological tool compounds have accordingly greatly contributed to elucidating the role of these channels in EDH-type dilator responses and other physiological functions. Since KCa channel pharmacology has previously been reviewed by us in great detail and put into its historical perspective⁵⁰, we will here only briefly mention the most commonly used modulators (Figure 3) and reserve more detailed discussions to compounds identified in the last three years.

FIGURE 3.

Structures and potencies of commonly used or recently developed KCa3.1 and KCa2 modulators.

Classical blockers

The “classical” blockers for KCa channels are, of course, charybdotoxin for KCa3.1 and apamin for the KCa2 channels⁵⁰. The larger charybdotoxin is a typical α-KTX scorpion toxin that anchors itself by two salt-bridges and inserts a central lysine residue into the selectivity filter of KCa3.1 and thus occludes the channel’s outer vestibule⁵¹. However, charybdotoxin never was an ideal KCa3.1 blocker because it also inhibits KCa1.1 (BK) and the voltage-gated Kv1.3 channels, both cross-reactivities, which initially caused confusion concerning the role of KCa3.1 in T cells and in vascular endothelium. Since the description of the clotrimazole derivatives TRAM-34⁵² and ICA-17043⁵³, which entered clinical trials for sickle cell anemia under the name senicapoc⁵⁴, charybdotoxin has been largely replaced as a pharmacological tool compound by these more KCa3.1 selective small molecule blockers. A new, structurally different KCa3.1 blocker is the benzothiazinone NS6180, which was recently identified in a Ca²⁺-activated Tl⁺ influx based high-throughput screen at NeuroSearch A/S in Denmark⁵⁵. NS6180 blocks KCa3.1 with an IC₅₀ of 11 nM by binding to the same two residues, Thr250 and Val275, in the inner pore as TRAM-34 and clotrimazole⁵⁶. Up to a concentration of 1 µM NS6180 and TRAM-34 are very selective KCa3.1 blockers and show no effect on the T-cell Ca²⁺ channel CRAC consisting of Orai-1 and STIM1 and a range of Kv, KCa, Nav, and TRP channels⁵⁵. However, at 10 µM both compounds start to significantly affect other channels. While NS6180 inhibits KCa1.1 (BK), Kv1.3 and Kv11.1 (hERG) by more than 50%, TRAM-34 blocks K_v1.3, K_v1.4, K_v7.2+K_v7.3 and Na_v1.4 at 10 µM⁵⁵. Both TRAM-34 and the new NS6180 should therefore not be used at concentrations above 1 µM, especially since both compounds also have very limited solubility around 10 µM and tend to quickly precipitate in aqueous solutions.

As mentioned above, the most widely used KCa2 blocker is the 18-amino acid honey bee venom peptide apamin⁵⁰, which is a remarkably selective blocker of KCa2 channels, with KCa2.2 being the most sensitive channel (IC₅₀ ~200 pM) and KCa2.1 and KCa2.3 being slightly less sensitive (see Table in Figure 3). Although initially assumed to inhibit KCa2 channels by simply occluding the pore similar to the larger scorpion and snake toxins, apamin more recently has been shown to inhibit KCa2 currents via an allosteric mechanism involving an outer pore histidine⁵⁷ and residues in the S3-S4 extracellular loop⁵⁸. The same type of more allosteric blocking mechanism probably applies to the permanently charged bis-quinolinium cyclophane UCL 1684⁵⁹ and its derivatives, which were designed to mimic the charges of apamin, and which are also of remarkable potency and selectivity (for a review see⁵⁰). A less potent but highly KCa2.2 selective blocker is Lei-Dab⁷, a derivative of the scorpion toxin leiurotoxin I, in which one residue in a crucial signature motif is replaced by the unnatural amino acid diaminobutonoic acid⁶⁰.

Positive and negative gating modulators

In addition to these “straight” blockers, which are either binding in the outer pore like the peptide toxins or in the inner pore like the TRAMs and NS6180, KCa2/3 channels also possess positive and negative gating-modulators (Figure 3), which left-shift or right-shift the calcium-response curve of these calcium/calmodulin gated channels and apparently render the channels more or less calcium-sensitive. Interestingly, gating-modulation of KCa3.1 and KCa2 channels seems to be possible at several points along the chain of molecular events leading from Ca²⁺ binding to the calmodulin at the C-terminus to eventual opening of the physical gate of these channels, which seems to be deeply buried in the inner pore vestibule, close to or even overlapping with the K⁺ selectivity filter^61–64. While benzimidazolone-type KCa activators like EBIO, the structurally related oxime NS309⁶⁵ or the benzothiazole SKA-31⁶⁶, do not differentiate well between KCa2 and KCa3.1 channels, which is in keeping with their binding site being located in the C-terminal region, close to or at the calmodulin binding domain²², the negative KCa2 channel gating-modulator NS8593 and the KCa2.1 selective (−)CM-TMPF have been found to interact with positions deep within the inner pore vestibule^67,68, where the gate of these channels is located and accordingly exhibit subtype selectivity. The negative gating modulator NS8593, which interacts with Ser507 in the pore-loop near the selectivity filter and A532 in an adjacent position in S6, exerts no effect on KCa3.1 channels but inhibits all three KCa2 channels at submicromolar concentrations⁶⁷. The very recently described [1,2,4]triazolo[1,5-a]pyrimidine (−)CM-TMPF and the structurally related (−)B-TMPF interact with Ser293 in S5 and act as KCa2.1 selective negative and positive gating modulators with IC₅₀ or EC₅₀ values of 24 and 31 nM⁶⁸. Both compounds are 40 to 100 times more potent than their respective (+)-enantiomers and display 10–20-fold selectivity over KCa2.2, KCa2.3 and KCa3.1, properties which make them valuable tool compounds for probing the currently not well defined physiological or pathophysiological role of KCa2.1. Another new and probably in the future very useful KCa2 channel modulator is the CyPPA derivative NS13001, which activates KCa2.2 and KCa2.3 with EC₅₀ values of 2 µM and 140 nM, but has no effects on KCa2.1 and KCa3.1⁶⁹.

Of these compounds, the ones that have so far most commonly been used in vivo are apamin (which interestingly is able to cross the blood-brain barrier and induces seizures following intraperitoneal application⁷⁰) and NS8593 as KCa2 channel inhibitors, TRAM-34 and senicapoc as KCa3.1 blockers, and SKA-31 as a mixed KCa2/3 channel activator, with ~10-fold selectivity for KCa3.1⁶⁶. Despite its high potency, NS309 is unfortunately not suitable for in vivo use due to its extremely short half-life and its 1 µM IC₅₀ for Kv11.1 (hERG)⁶⁵.

CLINICAL AND PRECLINICAL EXPERIENCE WITH KCa3.1 BLOCKERS

Past experiences

Apart from a single nucleotide polymorphism associated with ileal Crohn’s disease in an Australian population⁷¹, KCa3.1 mutations have so far not been described to be involved in human diseases. Nevertheless, KCa3.1 blockade constitutes a relatively-well validated therapeutic approach for immunosuppression and for curbing vascular smooth muscle cell and fibroblast proliferation¹². The oldest indication for KCa3.1 blockers is prevention of erythrocyte dehydration in sickle cell disease through inhibition of the so-called ‘Gàrdos channel’, the erythrocyte KCa3.1 channel. Early proof-of-concept studies from Carlo Brugnara’s group at the Children’s Hospital in Boston demonstrated that the unselective KCa3.1 blocker clotrimazole reduced erythrocyte dehydration in a transgenic mouse model⁷² and in 5 patients with sickle cell disease⁷³. Following up on this, Icagen Inc. advanced the more selective senicapoc⁵³ into clinical trials and reported that the compound significantly reduced hemolysis and increased hemoglobin levels in a 12-week, multicenter, randomized double-blind Phase-2 study⁵⁴. However, in a subsequent Phase-3 study, which was designed to compare the rate of acute vaso-occlusive pain crisis occurring in sickle cell disease patients, senicapoc failed to reduce this desired clinical endpoint and was terminated early, despite patients in the senicapoc group again showing improvements in anaemia and haemolysis⁵⁴. Since senicapoc was safe and well tolerated in these studies, Icagen Inc. next explored asthma as a potential therapeutic indication based on a large body of work demonstrating that KCa3.1 is expressed in various cell types involved in the pathogenesis of asthma (mast cell, macrophages, fibroblasts, airway epithelium and airway smooth muscle cells)⁷⁴ and that KCa3.1 blockade inhibits mast cell degranulation and migration^75,76 as well as airway smooth muscle proliferation^9,77. Following an initial evaluation in allergen induced asthma in sheep⁷⁸, senicapoc demonstrated encouraging results in a small Phase-2 allergen challenge study in patients with allergic asthma. However, in a second Phase-2 trial in exercise induced asthma the compound failed to demonstrate any significant improvement in lung function¹².

KCa3.1 blockade and/or disruption of the KCa3.1 gene has further been found to ameliorate various autoimmune and cardiovascular disease models through a combination of inhibiting immune cell proliferation, infiltration and cytokine production as well as dampening smooth muscle cell, fibroblast and vascular endothelial cell proliferation. For example, TRAM-34 or senicapoc prevent MOG induced autoimmune encephalomyelitis⁷⁹, anti-collagen antibody-induced arthritis⁸⁰, and trinitrobenzene sulfonic acid-induced colitis in mice⁸¹, renal fibrosis following unilateral uretral obstruction in mice and rats⁸², angiogenesis in the mouse matrigel plug assay⁸³, atherosclerosis development in ApoE^−/− mice⁸⁴, as well as angioplasty induced intimal smooth muscle hyperplasia (restenosis) in rats and pigs^9,85. KCa3.1 blockade has further been found to reduce microglia activation⁸⁶ and thus curb inflammatory responses and reduce neuronal damage in models of ischemic stroke⁸⁷, traumatic brain injury⁸⁸, optic nerve transection⁸⁹, and traumatic spinal cord injury⁹⁰.

Future potential

Despite the so far disappointing clinical trial results with KCa3.1 blockers, KCa3.1 remains an attractive pharmacological target for a variety of indications including postangioplasty restenosis, atherosclerosis, inflammatory bowel disease, and possibly neuroinflammation in the context of stroke, multiple sclerosis and Alzheimer’s disease⁹¹. Even asthma should probably not yet be dismissed as an indication since the performed studies were certainly not long enough to determine whether KCa3.1 blockers can prevent airway remodeling as would be expected from their inhibitory effects on airway smooth muscle and fibroblast proliferation⁷⁴. However, given the important role that KCa3.1 channels play in EDH responses, proposing to develop KCa3.1 blockers for any of the above named indications of course raises the question of whether KCa3.1 blockers will increase blood pressure, particularly when considering the higher blood pressure in the KCa3.1 and/or KCa2.3-deficient mice as outlined above. Pharmacological inhibition of KCa3.1, in contrast, has not been observed to raise blood pressure in mice⁸⁴ or in over 500 human volunteers and patients taking senicapoc for up to two years⁹², suggesting that combined blockade of endothelial KCa3.1 and KCa2.3 channels might be necessary to significantly raise blood pressure in humans. These findings, of course, do not discount that KCa3.1 blockade might not lead to increases in blood pressure and associated cardiovascular risk in situations like diabetes, hypertension and obesity, where EDH responses are known to be compromised^39,93. Moreover, KCa3.1 blockers may worsen the vascular dementia and cerebrovascular ischemia in aged patients. It, however, should be kept in mind that the beneficial anti-inflammatory and anti-atherosclerotic effects of KCa3.1 blockers may outweigh the impairments in endothelial responses and thus decrease cardiovascular risk overall. Taken together, the potential therapeutic usage of KCa3.1 blockers should therefore be carefully considered on the basis of a patient’s individual cardiovascular risk profile, the intended indication and side-effect profiles of the therapeutic alternatives. For example, in situations where the alternatives are cyclosporine or glucocorticoids, both medications that are diabetogenic and induce hypertension and/or kidney toxicity, a KCa3.1 blocker might be viewed very favorably. However, it will require longer-term animal studies and ideally clinical trials to ultimately assess cardiovascular risks of KCa3.1 blockers.

IN VIVO EXPERIENCE WITH KCa2 MODULATORS

As described above, KCa2 channels play an important role in controlling neuronal excitability and firing frequency by determining the magnitude of the mAHP. Pharmacological modulation of KCa channels, therefore, offers the opportunity to significantly affect neuronal activity. While KCa2 channel blockers like apamin induce high frequency bursts in neurons that normally exhibit regular tonic action potential firing³⁰ and elicit seizures in rodents⁹⁴, KCa2 channel activators like EBIO reduce neuronal firing rates and suppress epileptiform activity in cultured hippocampal neurons^22,23,95, suggesting that KCa2 channel activators might be useful for the treatment of CNS disorders that are characterized by hyperexcitability such as epilepsy and ataxia⁵⁰. This therapeutic hypothesis is supported by the finding that transgenic mice overexpressing SK3-1B, a dominant negative alternative transcript of KCa2.3, which suppresses KCa2 channels, exhibit severe ataxia with in-coordination, tremor and altered gate due to increased excitability of their deep cerebellar neurons⁹⁶. Conversely, acute KCa2 channel activation with SKA-31 partially corrects abnormal Purkinje cell firing and improves motor function in SCA3 mice⁹⁷, a model of spinocerebellar ataxia type 3, while 3 weeks of oral administration of NS13001 has recently been shown to improve performance of aging SCA2 mice (a model of spinocerebellar ataxia type 2) in motor coordination assays and to reduce Purkinje cell degeneration⁶⁹. Additional evidence supporting KCa2 channels as targets for ataxia comes from the observation that direct infusion of EBIO into the cerebellum significantly improves the motor coordination deficits in ducky and the dyskinesia and ataxia in tottering mice⁹⁸, both harboring mutations in P/Q-type Cav channels leading to loss of precision in Purkinje cell firing. EBIO has further been reported to reduce electroshock and pentylenetetrazole induced seizures⁹⁹ demonstrating that KCa2 activators are indeed anticonvulsants in vivo, a finding that has been recently confirmed by our own laboratory with SKA-31 in collaboration with the NIH Anticonvulsant Screening program (unpublished results). While potential blood pressure lowering effects of KCa2 activators, especially if KCa2.2 selective compounds are developed, are less of a concern for ataxia and epilepsy treatment, there is the possibility that KCa2 channel activators might affect learning and memory since KCa2 channels play an important role in synaptic plasticity and long-term potentiation (LTP)^18,100. Apamin administration in rats or mice enhance performance in the Morris water maze, the object recognition test and the appetitive lever-pressing task^101–104, while mice overexpressing KCa2.2 exhibit impaired hippocampal-dependent learning and memory in both the Morris water maze and a contextual fear-conditioning paradigm¹⁰⁵. Similarly, pharmacological KCa2 channel activation with EBIO or CyPPA has been reported to impair the encoding, but not the retrieval of object memory in a spontaneous object recognition task¹⁰⁶. However, in this context it should also be mentioned that targeted overexpression of KCa2.2 in the basolateral amygdala has been found to reduce anxiety and stress-induced corticosterone secretion¹⁰⁷ and that EBIO and CyPPA administration has recently been reported to reduce fear memory caused by repeated stress in rats, suggesting the use of KCa2 activators for alleviating psychiatric symptoms associated with stress induced amygdala hyperactivity¹⁰⁸. There is also increasing evidence that neuronal KCa2 channel expression is highly plastic. For example, hippocampal KCa2.3 expression increases with age¹⁰⁰, while decreased KCa2.3 expression in the nucleus accumbens is associated with alcohol dependence and infusion of EBIO into the nucleus accumbens reduces alcohol seeking behavior in alcohol dependent rats¹⁰⁹. In contrast, pilocarpine-treated chronically epileptic rats exhibit decreased KCa2.1 and KCa2.2 expression in the hippocampus¹¹⁰ and reduced afterhyperpolarization in CA1 hippocampal neurons¹¹¹.

In addition to reducing neuronal excitability, KCa2 activators might also be useful for treating erectile dysfunction, urinary incontinence and for reducing uterine contractions. Both overexpression of KCa2.3 and intravesicular instillation of NS309 improve bladder function^34,112, while CyPPA has recently been shown to inhibit phasic uterine contractions and delays preterm birth in mice¹¹³, a finding which is in keeping with the observations made in transgenic mice that KCa2.3 is involved in respiration and parturition¹¹⁴.

Furthermore, KCa2 channels are involved in cardiac excitability and are expressed at higher levels in the atria than the ventricles of mice, rats, rabbits and humans^35,115. KCa2 inhibitors have accordingly been proposed as potential new drugs to treat atrial fibrillation (AF)³⁷. As in other tissues, KCa2 channels in the heart seem to be tightly coupled to Ca²⁺ sources and KCa2.2 have been found to associate with Cav1.2 and Cav1.3 channels in atrial myocytes through the cytoskeletal protein α-actinin2¹¹⁶. Changes in KCa2 channel expression like the reported increase of KCa2.3 expression in failing rabbit ventricles¹¹⁷, which leads to action potential shortening, or the KCa2.3 variations found in lone AF³⁶, have been associated with ventricular or atrial fibrillation. In keeping with the idea that KCa2 blockers are antiarrhythmic, apamin increases action potential duration in human, mouse and rabbit atrial myocytes^35,117 and both the NS8593 and UCL 1684 have been reported to terminate acetylcholine-induced AF in rat, guinea pig and rabbit hearts¹¹⁸ and to reduce AF duration in normotensive and hypertensive rats¹¹⁹.

DO ACTVATORS OF ENDOTHELIAL KCa CHANNELS HAVE THERAPEUTIC POTENTIAL AS NOVEL ANTIHYPERTENSIVES?

There is no doubt that KCa3.1/KCa2 activators exert cardiovascular effects in vivo. In fact, i.p. injections of SKA-31 at concentrations of 10–30 mg/kg lower blood pressure in mice, a depressor response that is considerable and last for 60–90 minutes (Delta approx. −30 mmHg)¹²⁰. Blood pressure lowering effects have been further documented in murine models of hypertension like angiotensin-II induced hypertension⁶⁶ or in transgenic mice expressing human angiotensinogen and renin (hAGN/hR)¹²¹. SKA-31 is also active in large mammals with a heart rate and blood pressure more similar to humans. In instrumented, conscious dogs i.v. injection of 2 mg/kg SKA-31 or of its derivative SKA-20⁶⁶ produced an immediate and strong (−30 mmHg), but short-lived (10–20 sec) depressor response¹²². These studies in dogs and mice showed that KCa3.1 and KCa2.3 channels are principally available as antihypertensive drug targets. Moreover, the experiments demonstrated for the first time that the meanwhile “classical” EDH-system can be targeted to lower blood pressure in vivo and could thus constitute an alternative to other antihypertensives like Ca²⁺ channel antagonist or NO-donors acting directly on the smooth muscle.

Unresolved issues

However, there are several unresolved issues that need to be addressed before it is possible to determine whether KCa3.1 and KCa2.3 activators will be clinically viable and useful antihypertensives. 1) We currently do not know whether continuous stimulation of endothelial KCa2/3 channels will result in desensitization either through “feed-back” modulation of Ca²⁺ sensitivity and channel activity or reduction in channel expression. That the latter phenomenon is a possibility has been suggested by a study on keratinocytes where EBIO treatment unexpectedly down-regulated KCa3.1 expression and inhibited proliferation¹²³. 2) Sedation and possible impairment of learning and memory through activation of neuronal KCa2 channels could constitute a serious enough side-effect to prohibit the use of KCa2 activators for treating hypertension. Indeed, we recently showed that i.p. injections of SKA-31 at 30 mg/kg produce substantial and lasting sedation (over 1 hour) in mice and 10 and 30 mg/kg SKA-31 reduce voluntary home cage activity over 24 hours¹²⁴. This problem could of course be avoided by either developing KCa3.1 selective activators or mixed KCa2.3/KCa3.1 activators that are not brain penetrant. 3) KCa2 activators could potentially induce cardiac side effects by causing blockade of AV-transmission or slowing down heart frequency through prolongation of atrial and/or ventricular APs. Indeed, at 30 mg/kg SKA-31 produces pronounced bradycardia in both wild-type and KCa3.1-deficient mice¹²⁰ suggesting that the effect is most likely caused by activation of cardiac KCa2.X channels. On the other hand, acute i.v. administration of SKA-31 in dogs produces reflex tachycardia to the depressor response¹²². The latter indicates that KCa3.1/KCa2.3 activators elicit very species-specific cardiovascular effects, which in this particular case are most probably related to the different heart rates of mice and dogs and their different cardiovascular regulatory mechanisms. 4) Another principal concern is that endothelial vasodilator function is often compromised in patients with essential hypertension, obesity or diabetes^39,93 and presumably also neurodegenerative disorders like Alzheimer’s disease, and that the endothelial KCa channels might not be sufficiently available as targets or their activation not effective enough. 5) Lastly, there is the possibility that KCa3.1 activators could stimulate immune cell functions like migration or cytokine and reactive oxygen species production and thus exacerbate inflammation. However, we think that this is an unlikely scenario since lymphocytes will probably counter-regulate. We think it equally improbable that KCa3.1 activators would stimulate hyperplasia and fibrosis. Nevertheless, these possibilities should be experimentally investigated before KCa3.1 activators could enter clinical development.

CONCLUSIONS

Considering all these potential caveats, we think it unlikely that a major pharmaceutical company would, at the current stage, invest the resources to develop KCa3.1/KCa2.3 activators for long-term treatment of essential hypertension given the availability of many potent and highly effective antihypertensives such as angiotensin converting enzyme inhibitors or angiotensin receptor blockers. However, KCa activators, especially KCa3.1 selective compounds or mixed KCa3.1/KCa2.3 activators that do not enter the brain, certainly have therapeutic potential for short term applications like intra-surgical hypertension, acute vasospasm or local application after stent implantations. One especially interesting application might be preservation of endothelial function in large vascular organs like hearts or kidneys or in vessel grafts during storage and transplantation. For this usage KCa3.1/KCa2.3 activators could simply be included in the cardioplegia solution. KCa3.1 activators might further be useful for combination therapy in situations where existing antihypertensives are not sufficiently effective or when it is considered desirable to increase blood flow in the microcirculation.