Cardiac and Renal Inward Rectifier Potassium Channel Pharmacology: Emerging Tools for Integrative Physiology and Therapeutics

Abstract

Inward rectifier potassium (Kir) channels play fundamental roles in cardiac and renal function and may represent unexploited drug targets for cardiovascular diseases. However, the limited pharmacology of Kir channels has slowed progress toward exploring their integrative physiology and therapeutic potential. Here, we review recent progress toward developing the small-molecule pharmacology for Kir2.x, Kir4.1, and Kir7.1 and discuss common mechanistic themes that may help guide future Kir channel-directed drug discovery efforts.

Introduction

Inward rectifier potassium (Kir) channels play key roles in cardiac excitation-contraction coupling, renal water and solute transport, and other vital physiological and pathophysiological processes [1]. In mammals, the channel superfamily is comprised of at least 16 genes (KCNJx) and 7 sub-families (Kir1.x–7.x) that share a common molecular structure [2]. Kir channels are tetramers of identical (homomeric) or similar (heteromeric) subunits assembled around an aqueous membrane-spanning pore. They lack regulatory voltage-sensing domains, but are gated by polyvalent cations (e.g. polyamines and Mg²⁺ [3–5]) that occlude the pore at cell potentials more positive than the K⁺ equilibrium potential (E_K). Kir channels thus function as biological diodes by limiting the extent of outward, but not inward, K⁺ current, a property is termed inward rectification. Strong rectifiers exhibit a sharp cutoff of outward current due to the presence of negatively charged pore-lining residues that stabilize electrostatic interactions with pore-blocking cations (Fig. 1), whereas weak rectifiers, have uncharged amino acids at some of the equivalent positions that weaken block and enable the channels to remain open well above E_K [1].

Figure 1. Putative small-molecule binding sites in a model Kir channel.

A) Side B) Top and C) Bottom view of chicken Kir2.2 crystal structure with equivalent residues implicated in small-molecule binding highlighted with sphere. Channel-specific residues are indicated in the legend. See text and Table 1 for details.

Kir channels are also gated through interactions with the membrane phospholipid phosphatidylinositol 4,5-bisphosphate (PIP₂). X-ray structures of Kir2.2 revealed that binding of a PIP₂-derivative induces conformational changes in the cytoplasmic domain (Fig. 1) that opens the pore [6]. Growing evidence suggests that some cationic amphiphilic drugs (e.g. carvedilol, mefloquone, thiopental) inhibit Kir channels by interfering with channel-PIP₂ interactions [7–9]. These have been discussed recently elsewhere [10] and will not be considered further.

As discussed below, emerging physiological, genetic, and pharmacological data point to specific Kir channels as novel drug targets for cardiac and renal diseases. With few exceptions, the rudimentary pharmacology has mired efforts to explore their integrative physiology and therapeutic potential of Kir channels. This review provides a snapshot of the current state of the field, highlighting recent progress and opportunities for developing the pharmacology of cardiac and renal Kir channels.

Cardiac Kir2.x channels

The ability of the heart to function as a pump requires that atrial and ventricular chambers contract in a highly stereotyped and synchronized fashion. Action potentials (AP) originating in the sinoatrial node (SAN) spread through the atria and then ventricles to initiate contractions. Three Kir channel sub-families contribute to cardiac excitability. Heteromeric Kir3.1/3.4 channels comprising the muscarinic M2 receptor-activated I_KACh current slows SAN pacemaker discharge and heart rate in response to parasympathetic nerve stimulation. Activation I_KATP carried by heteromeric Kir6.2/SUR channels during metabolic stress contributes to ischemic preconditioning that protects heart function during prolonged ischemia. The molecular physiology, pathophysiology, and pharmacology of I_KACh and I_KATP have been reviewed extensively [1,11–14] and will not be discussed here. The major focus of this review is on the molecular pharmacology of the third group of cardiac Kir channels: the strong rectifiers Kir2.1, Kir2.2, and Kir2.3. Homomeric and heteromeric assemblies of Kir2.x subunits underlie the I_K1 current that dominates the resting K⁺ conductance and shapes late-phase action potential (AP) repolarization in cardiac myocytes [12]. Genetic loss- and gain-of-function mutations in Kir2.1 (KCNJ2) prolong and shorten, respectively, the AP duration and increase the susceptibility to lethal ventricular arrhythmias [15,16]. No disease-causing mutations in Kir2.2 (KCNJ12) and Kir2.3 (KCNJ4) have been reported.

It is clear that Kir2.x channels play important roles in heart pump function, but their species and regional heterogeneity and lack of specific pharmacological probes has slowed efforts to develop a comprehensive understanding of their integrative physiology and druggability in cardiac diseases. Below we discuss recent progress toward developing these critically needed tools for overcoming this barrier.

Chloroquine

The 4-aminoquinoline derivative chloroquine (Table 1) is used widely as an anti-malarial drug in developing countries. However, prolonged treatment or overdose can induce lethal ventricular arrhythmias through inhibition of various cardiac ion channels [17]. Sanchez-Chapula and colleagues [18] found that chloroquine blocks Kir2.1 at clinically relevant doses (half-maximal inhibitor concentration [IC₅₀] = 8.7 μM) and in a voltage-dependent manner consistent with ‘knock-off’ of the drug from the intracellular pore [19]. Direct application of chloroquine to the cytoplasmic face of Kir2.1 results in channel inhibition that is much faster (~15 sec) than that observed when applied extracellularly (~8 min), suggesting chloroquine must cross the plasma membrane to reach an intracellular binding site. Indeed, alanine-scanning mutagenesis revealed that mutation of four residues in the cytoplasmic domain of Kir2.1 led to progressive loss of chloroquine sensitivity with the following rank-order: E224 > D259 > E299 > F254 (Fig. 1). Molecular modeling identified an energetically favorable docking pose between chloroquine and the channel involving electrostatic interactions between E224, D259, and E299 as well as aromatic pi-stacking with P254. Identifying the putative chloroquine binding site creates opportunities for designing safer analogs exhibiting reduced Kir2.1 activity and cardiotoxicity.

Table 1. Chemical diversity of Kir channel modulators.

Compound/PubChem CID	Structure	Action	Selectivity (IC50/EC50)*	Channel/Residues	References
Chloroquine/2719		Pore blocker	IKACh (0.4) > IKATP (0.5) > IK1 (0.7)	Kir2.1/E224, D259, E299, F254	[18,22]
Pentamidine/4735		Pore blocker	Kir2.1 (0.2) > 2.2 > Kir2.3	Kir2.1/E224, D259, E299	[25]
ML133/781301		Pore blocker	Kir2.1 (0.3†, 1.9) > 2.6 (2.8) > 2.2 (2.9) > 2.3 (4.0) > 6.2 (7.7) > 7.1 (32.9) > 4.1 (76) > 1.1 (>300)	Kir2.1/D172, I176	[28]
Flecainide/3356		Potentiator	Kir2.1 (0.4†† − 0.8‡) > 2.2 ~ 2.3 ~ 2.1/2.2 ~ 2.1/2.3	Kir2.1/E224, D259, E299, C311	[33]
Fluoxetine/62857		Pore blocker	Kir4.1 (15) > 2.1 ~ 1.1	Kir4.1/T128, L151, E158	[46,48]
Nortriptyline/4543		Pore blocker	Kir4.1 (28) > 2.1 (245)	Kir4.1/T128, L151, E158	[47,48]
VU590/4536383		Pore blocker	Kir1.1 (0.3) > 7.1 (8) > 2.1 ~ 4.1	Unknown	[39]
VU573/44199324		Unknown	Kir3.1/3.2 (2) > 2.3 (5) > 7.1 (5)	Unknown	[52]

See text for details.

^*μM,

^†at pH 8.5;

^††at 50 mV,

^‡at −120 mV

Pharmacological Kir2.1inhibition may provide therapeutic benefits in the setting of certain cardiac pathologies. Gain-of-function mutations in KCNJ2 underlying Short QT Syndrome-Type 3 (SQT3) increase the risk of lethal atrial and ventricular arrhythmias [15]. These mutations (D172N, E299V, and M301K) reduce pore block by Mg²⁺/polyamines and produce larger outward currents that shorten AP duration. Heterologous expression and in silico studies have suggested that inhibition of the SQT3 mutant Kir2.1-D172N by chloroquine may normalized the AP waveform and improve cardiac function in SQT3 patients [20,21]. This may not bet true for patients carrying E299V and M301K mutations since these residues are near the putative chloroquine binding site (Table 1).

Kir2.1 inhibition may also restore sinus rhythm in the setting chronic atrial fibrillation, where upregulation of Kir2.1 expression and function contributes to arrhythmia recurrence and maintenance. Jalife and colleagues demonstrated that chloroquine treatment terminates atrial or ventricular fibrillation in mice, rabbit, and sheep hearts. At the doses used, however, chloroquine also inhibits I_KACh and I_KATP [22]. Additional studies are needed to determine if specific Kir2.1 inhibition is sufficient to restore rhythmogenesis.

Pentamidine

Pentamidine is an aromatic diamidine (Table 1) used in the treatment of pneumocystis pneumonia, trypanosomiasis, leishmaniasis, and fungal infections. Intravenous administration in some patients delays ventricular repolarization and induces ventricular tachycardia, consistent with hERG block. Although pentamidine does indeed inhibit hERG function, this requires prolonged drug treatment and is mediated through inhibition of channel trafficking [23], a mechanism that is too slow to explain the more rapid cardiac effects of the drug [24]. Van der Hayden and colleagues [25] reasoned that some of the effects could be due to inhibition of I_K1 current carried by members of the Kir2.x subfamily. Indeed, they found that acute pentamidine application at clinically relevant doses inhibits homotetrameric Kir2.x channels with the rank-order potency Kir2.1 (IC₅₀ = 190 nM) > Kir2.2 > Kir2.3. Similar to chloroquine (see above), pentamidine inhibited Kir2.1 much more rapidly when applied to the cytoplasmic side of the channel. In silico ligand docking into the Kir2.1 cytoplasmic domain crystal structure [26] identified a putative pentamidine binding site involving E224, D259, and E299 (Fig. 1), the same residues implicated in chloroquine block (Table 1). Mutations of these residues reduced pentamidine sensitivity, lending support to their model predictions.

In an effort to develop more specific Kir2.1 probes, the investigators analyzed a series of pentamidine analogs for activity toward Kir2.x and several other cardiac ion channels [27]. One compound termed PA-6 exhibited ~15 nM affinity toward Kir2.1, Kir2.2, and Kir2.3, but no discernible activity toward cardiac sodium, calcium, and potassium currents. In terms of potency, PA-6 represents the state-of-the-art in Kir2.1 inhibitors. PA-6 was active on native I_K1 in cardiomyocytes and predictably lengthened the action potential duration. Interestingly, and unlike pentamidine, PA-6 had no effect on the hERG biosynthesis. Provided the pharmacokinetic behavior is amenable, PA-6 should provide a useful in vivo probe of Kir2.x function. Furthermore, and given the apparent lack of effect on hERG, PA-6 may provide a safer alternative to pentamidine for the treatment protozoal infections.

ML133

Min Li and colleagues took a modern drug discovery approach to develop a potent and selective small-molecule inhibitor of Kir2.1 [28]. Using a fluorescence-based thallium (Tl⁺) flux assay [29,30], the investigators screened approximately 300,000 structurally diverse small molecules from the NIH Molecular Libraries Small-Molecule Repository for chemical modulators of Kir2.1. Of 320 confirmed hits, ML133 (N-(4-methoxybenzyl)-1-(naphthalene-1-yl)methanamine) was selected for development because of its drug-like structure, potency toward Kir2.1, and clean ancillary pharmacology. At physiological pH, ML133 exhibits the following rank-order potency: Kir2.1 (IC₅₀ = 2 μM) > 2.6 > 2.2 > 2.3 > 6.2 > 7.1 > 4.1 > >1.1 (Table 1). Interestingly, due to the presence of a protonatable nitrogen in the linker of ML133, its potency toward Kir2.1 is pH-dependent: IC₅₀ = 300 nM at pH 8.5 and 10 μM at pH 6.5 (Table 1). Lead-optimization efforts to improve its potency were unsuccessful. Similar to chloroquine (see above), block of Kir2.1 by ML133 was dependent on the K⁺ electrochemical driving force, suggesting the binding site is located in ion conduction pathway. In an effort to identify the binding site, the investigators took advantage of the >50-fold selectivity of ML133 for Kir2.1 over Kir1.1 and analyzed a series of Kir1.1-Kir2.1 chimeras and point mutants for sensitivity to ML133. Swapping out incrementally larger regions of Kir2.1 for Kir1.1 localized a potential binding site to the Kir2.1 pore. This region of Kir2.1 and Kir1.1 is highly conserved between the two channels, with the exception of two conspicuous residues: D172 and I176 (Fig. 1). Mutation of these residues to those found in Kir1.1 either alone (i.e. Kir2.1-D172N) or together (Kir2.1-D172N/I176C) reduces ML133 potency by 8- and 23-fold, whereas the reverse mutations in Kir1.1 (i.e. Kir1.1-N171D and Kir1.1-N171D/C175I) confers weak sensitivity to the channel. This putative binding site is different from that of chloroquine and pentamidine (Fig. 1). In terms of selectivity, and based on published data, ML133 represents the state-of-the-art among Kir2.1 channel probes. However, similar to chloroquine and pentamidine, ML133 is not selective for members of the Kir2.x. A future challenge will be designing compounds that can be used to dissect the individual functions and druggability of Kir2.x homo- and heteromeric channels in the heart and other organs.

Flecainide

Flecainide is used widely as a class Ic antiarrhythmic drug that, in addition to inhibiting cardiac sodium channels, displays class III antiarrhythmic activity due to inhibition of Kv4.3 and hERG [31,32]. In an elegant study, Caballero and colleagues reported [33] that clinical doses of flecainide increase ventricular I_K1 through potentiation of Kir2.1, but not of Kir2.2, Kir2.3, Kir2.1/2.2 or Kir2.1/2.3 channels. Molecular modeling, mutagenesis, and patch clamp electrophysiology supports a mechanism in which flecainide binding to C311 of Kir2.1 reduces spermine-dependent pore block and increases outward current amplitude. Remarkably, prolonged flecainide treatment partially rescues the function of Kir2.1 channels carrying an R67W LOF mutation identified in Andersen syndrome (AS) patients by increasing the cell surface density of the channel. It is tempting to speculate that this mechanism underlies the propensity of flecainide to suppress ventricular arrhythmias in some AS patients [34,35].

Renal Kir channels

Of the nearly 70 million Americans with hypertension, an insidious disorder which increases the risk of heart disease, stroke, and kidney damage, only about half manage their blood pressure correctly. Diuretics represent the first-line therapy for hypertension and work by reducing sodium and osmotically obliged water reabsorption in the kidney tubule, leading to increased urine output and lowering of blood volume and pressure. The development of new structural classes of diuretic drugs with different mechanisms of action would provide clinicians with greater therapeutic options for managing their patients’ blood pressure, particularly in the settings of diuretic resistance and harmful drug-drug interactions [36]. Below, we discuss recent progress in developing the molecular pharmacology of renal Kir channels that could represent targets for novel-mechanism diuretics.

Kir1.1 (ROMK, KCNJ1)

Genetic LOF mutations in KCNJ1 lower blood pressure by inhibiting renal sodium and water reabsorption [37,38], thus validating Kir1.1 as a diuretic target. Recent advances in the small-molecule pharmacology by our group [39,40] and Merck [41] are providing critically needed tools for exploring Kir1.1 as a therapeutic target for hypertension. This work has been reviewed recently by our group [42] as well as by Garcia and Kaczorowski (2013) in this issue of Current Opinion in Pharmacology and will not be discussed further.

Kir4.1 (KCNJ10)

Inactivating mutations in the gene encoding Kir4.1 (KCNJ10) give rise to SeSAME (or EAST) syndrome, a complex disease presenting with seizures, sensorineural deafness, ataxia, intellectual disability, and electrolyte imbalance [43,44]. The renal consequences of SeSAME syndrome are consistent with impaired NaCl reabsorption in the distal convoluted tubule (for review see [42]). More recently, Zaika et al. (2013) found that dopamine inhibits Na⁺ reabsorption in the cortical collecting duct (CCD) through inhibition of Kir4.1 homomeric and Kir4.1/5.1 heteromeric channels [45]. Taken together, these data raise the intriguing possibility that renal Kir4.1-containing channels represent novel diuretic targets.

The molecular pharmacology of Kir4.1 is poorly developed and consists of selective serotonin reuptake inhibitors (SSRI), such as fluoxetine (IC₅₀ = 15 μM, [46]) and tricyclic antidepressants (TCA), such nortriptyline (IC₅₀ = 28 μM; [47]). Within the Kir channel family, these drugs appear to be selective for Kir4.1 since they exhibit very limited activity toward Kir1.1 and Kir2.1 [46,47]. Both drugs exhibit voltage-dependent inhibition of Kir4.1, consistent with a pore-blocking mechanism. Kurachi and colleagues found through alanine-scanning mutagenesis that T128, L151, and E158 are required for fluoxetine- and nortriptyline-dependent inhibition of Kir4.1 [48]. Mutation to D or E of N171 in Kir1.1 (N171D/E), which is equivalent to E158 in Kir4.1, confers fluoxetine and nortriptyline sensitivity to Kir1.1. Molecular docking simulations suggested that both drugs can assume energetically favorable interactions with E158 and T128. L151 is located outside of the pore and likely affects drug binding through stabilization of channel tertiary structure. Interestingly, their modeling suggests that both drugs interact with E158 and T128 on diagonally apposed subunits of the tetramer. This raises the possibility that SSRIs and TCAs are specific for Kir4.1 homomeric vs. Kir4.1/5.1 heteromeric channels since Kir5.1 contains an asparagine residue (N161) at the equivalent position of Kir4.1-E158. The development of small-molecule probes targeting Kir4.1 and Kir4.1/5.1 channels would create exciting opportunities for exploring the therapeutic potential of these channels in hypertension and other disorders.

Kir7.1 (KCNJ13)

As one of the newest members of the Kir channel family, there is little known about the function Kir7.1 in the organs known to express the channel, including eye, intestine, stomach, thyroid, spinal cord, brain, and kidney. Studies of Kir7.1 in retinal pigmented epithelial (RPE) cells have begun to shed light on the physiology of the channel in the eye, where it participates in transepithelial ion and water transport required for vision. A LOF mutation in KCNJ13 is associated with Snowflake Vitreoretinal Degeneration (SVD), indicating Kir7.1 function is indeed essential in the eye [49].

The function of Kir7.1 in the kidney is currently a matter of speculation based on its localization and observation that Kir7.1 expression is regulated by serum K⁺ concentration [50,51]. In rat, Kir7.1 is expressed in the basolateral membrane of several nephron segments, including principal cells of the CCD, which plays fundamental roles in regulated Na⁺ reabsorption and K⁺ excretion. Transepithelial K⁺ excretion in the CCD is mediated by a two-step process, whereby serum K⁺ is initially pumped across the basolateral membrane by the Na⁺-K⁺-ATPase and then secreted into the renal tubule fluid via apical Kir1.1 channels. By hyperpolarizing V_m, K⁺ secretion enhances the electrochemical driving force favoring Na⁺ reabsorption across the apical membrane through the K⁺-sparing diuretic target ENaC (for Epithelial Na⁺ Channel). Reabsorbed Na⁺ is transported by the Na⁺-K⁺-ATPase across the basolateral membrane before returning to the blood. Although there is no experimental evidence in support of this model, some investigators have postulated that Kir7.1 recycles K⁺ across the basolateral membrane to maintain robust Na⁺-K⁺-ATPase activity needed for transepithelial K⁺ secretion and Na⁺ reabsorption. If this model is correct, a Kir7.1 antagonist should 1) slow Na⁺-K⁺-ATPase activity, 2) reduce apical K⁺ secretion, 3) depolarize the apical V_m, 4) reduce ENaC-mediated Na⁺ reabsorption, and 5) lower blood volume and pressure. In principle, Kir7.1 inhibitors acting in the CCD would mimic the effects of the K⁺ sparing diuretic amiloride, but may also reduce Na⁺ reabsorption in other nephron segments such as the DCT, where Kir7.1 function may overlap with that of Kir4.1 (see above). To our knowledge, no studies have reported renal tubule pathologies in SVD patients carrying a loss-of-function mutation in KCNJ13. This could indicate that Kir7.1 is dispensable for renal function or that the loss of channel function leads to subclinical changes in blood pressure.

The small-molecule pharmacology of Kir7.1 is in its infancy and limited to two inhibitors developed by our group (Table 1). VU590 is a 300 nM inhibitor of Kir1.1 that also inhibits Kir7.1 with an IC₅₀ of ~8 μM [39]. The other is the mixed Kir channel inhibitor VU573, which inhibits Kir7.1 with an IC₅₀ of ~5 μM [52]. In an effort to expand the pharmacology, we recently developed a Tl⁺ flux assay of Kir7.1 that takes advantage of a pore mutation (M125R) that increases the unitary K⁺ conductance of the channel by ~20-fold [53]. Unlike the WT channel, Kir7.1-M125R mediates robust Tl⁺ flux that can be measured in a 384-well plate format that is compatible with HTS [52]. We anticipate that this assay will enable the development of small-molecule tool compounds for exploring Kir7.1 physiology and druggability in the kidney tubule and other organ systems.

Conclusions

In conclusion, a growing body of physiological, genetic and pharmacological evidence has implicated some Kir channel as drug targets for cardiac and renal diseases. The last 5 years have seen more progress in the development of the Kir channel pharmacology than in the previous 20 years since the founding member was cloned [54,55]. The identification of clinically used drugs (e.g. chloroquine [18], pentamidine [25,27], flecainide [33], fluoxetine [46,56], and nortriptyline [47]; Table 1) exhibiting Kir channel-directed activity creates opportunities for engineering more potent and selective Kir channel modulators from these chemical scaffolds (e.g. [27]). Chemically unique Kir channel inhibitors have been discovered in publically funded, large-scale screens of diverse chemical libraries, the results of which are deposited in well-annotated databases (http://pubchem.ncbi.nlm.nih.gov/). Counterscreens of focused libraries containing modulators of one Kir channel have uncovered modulators of other Kir channel members [52,57]. Despite this important progress, there are many hurdles to overcome before a comprehensive pharmacological “toolkit” for cardiac and renal inward rectifiers will be realized. These include the development of 1) subtype-selective modulators of the Kir2.x family, including compounds that discriminate between homo- and heteromeric channels, 2) modulators of Kir4.1 homomeric and Ki4.1/5.1 channels, and 3) modulators of Kir7.1. It is noteworthy that most known Kir channel pore blockers studied to date appear to target residues in the membrane-spanning or cytoplasmic pore domain (Fig. 1, Table 1). The development of comparative molecular models of these pore regions could enable in silico screening of “virtual libraries” containing millions of compounds at a fraction of the time, labor, and cost.

Highlights.

• Some Kir channels are putative drug targets, but their pharmacology is limited

• Newly discovered inhibitors or activators of Kir2.x, 4.1, and 7.1 are discussed

• Mechanisms of action involve pore-lining residues that control rectification

• This work is a critical first step toward developing more specific modulators