Advertisement
Research ArticleArticle

Structural Determinants for the Selectivity of the Positive KCa3.1 Gating Modulator 5-Methylnaphtho[2,1-d]oxazol-2-amine (SKA-121)

Brandon M. Brown, Heesung Shim, Miao Zhang, Vladimir Yarov-Yarovoy and Heike Wulff
Molecular Pharmacology October 2017, 92 (4) 469-480; DOI: https://doi.org/10.1124/mol.117.109421

Abstract

Intermediate-conductance (KCa3.1) and small-conductance (KCa2) calcium-activated K+ channels are gated by calcium binding to calmodulin (CaM) molecules associated with the calmodulin-binding domain (CaM-BD) of these channels. The existing KCa activators, such as naphtho[1,2-d]thiazol-2-ylamine (SKA-31), 6,7-dichloro-1H-indole-2,3-dione 3-oxime (NS309), and 1-ethylbenzimidazolin-2-one (EBIO), activate both channel types with similar potencies. In a previous chemistry effort, we optimized the benzothiazole pharmacophore of SKA-31 toward KCa3.1 selectivity and identified 5-methylnaphtho[2,1-d]oxazol-2-amine (SKA-121), which exhibits 40-fold selectivity for KCa3.1 over KCa2.3. To understand why introduction of a single CH3 group in five-position of the benzothiazole/oxazole system could achieve such a gain in selectivity for KCa3.1 over KCa2.3, we first localized the binding site of the benzothiazoles/oxazoles to the CaM-BD/CaM interface and then used computational modeling software to generate models of the KCa3.1 and KCa2.3 CaM-BD/CaM complexes with SKA-121. Based on a combination of mutagenesis and structural modeling, we suggest that all benzothiazole/oxazole-type KCa activators bind relatively “deep” in the CaM-BD/CaM interface and hydrogen bond with E54 on CaM. In KCa3.1, SKA-121 forms an additional hydrogen bond network with R362. In contrast, NS309 sits more “forward” and directly hydrogen bonds with R362 in KCa3.1. Mutating R362 to serine, the corresponding residue in KCa2.3 reduces the potency of SKA-121 by 7-fold, suggesting that R362 is responsible for the generally greater potency of KCa activators on KCa3.1. The increase in SKA-121’s KCa3.1 selectivity compared with its parent, SKA-31, seems to be due to better overall shape complementarity and hydrophobic interactions with S372 and M368 on KCa3.1 and M72 on CaM at the KCa3.1–CaM-BD/CaM interface.

Introduction

Small- and intermediate-conductance calcium-activated potassium channels (KCa) are gated by the binding of calcium to calmodulin (CaM), which is constitutively associated with the calmodulin-binding domain (CaM-BD) on the C terminus of these voltage-independent channels. There are four channels in the so-called KCa2/3 group: the small-conductance KCa2.1, KCa2.2, and KCa2.3, known collectively as SK channels, and the intermediate-conductance KCa3.1 channel, also known as IK (Wei et al., 2005). KCa2/3 channels share a similar overall structure with the voltage-gated potassium channel family. The channels are tetramers consisting of four six-transmembrane domains with the N and C termini located on the cytoplasmic side of the membrane. KCa2 and KCa3.1 channels have the same Ca2+/calmodulin-mediated gating mechanism (Xia et al., 1998; Fanger et al., 1999) but differ in their conductance, with KCa2 channels having a single channel conductance of ∼2 pS and KCa3.1 having a single channel conductance of 11 pS at physiologic potassium concentrations (Wei et al., 2005). Whereas KCa2 channels are primarily expressed in the nervous system (Adelman et al., 2012), KCa3.1 is predominantly found in peripheral tissues (Wulff and Castle, 2010). In both cases, KCa channel activation hyperpolarizes the cell membrane, inducing different physiologic effects depending on the location of expression. For example, in immune cells, KCa3.1 channels regulate calcium influx and cellular activation (Feske et al., 2015); in the vascular endothelium, both KCa3.1 and KCa2.3 contribute to controlling vascular tone (Busse et al., 2002); and in neurons, KCa2 channels mediate afterhyperpolarization, regulate firing frequency, and contribute to learning and memory (Adelman et al., 2012). Therefore, pharmacologic activation of KCa channels seems attractive for the treatment of neurologic and cardiovascular diseases (Lam et al., 2013; Christophersen and Wulff, 2015). Specifically, KCa2 channel activation has been proposed for reducing neuronal excitability in epilepsy and ataxia, whereas KCa3.1 activation could potentially provide a novel therapeutic strategy to treat hypertension.

Although the therapeutic potential of activating KCa2/3 channels is clear, it has been a challenge to develop subtype selective compounds. The majority of KCa2/3 channel activators suffer from a lack of selectivity, displaying only a modest 5- to 10-fold selectivity for KCa3.1. In 1996, the first compound identified to activate KCa channels was 1-EBIO, or simply EBIO (1-ethylbenzimidazolin-2-one) (Devor et al., 1996). EBIO has been reported to activate KCa3.1 with EC50 values ranging from 28 to 210 μM (Jensen et al., 1998; Syme et al., 2000; Pedarzani et al., 2001; Singh et al., 2001), KCa2.1 at 630 μM, KCa2.2 at 500 μM to 1 mM, and KCa2.3 at 170 μM to 1 mM (Cao et al., 2001; Pedarzani et al., 2001; Lappin et al., 2005; Hougaard et al., 2007). Another widely used KCa2/3 activator is SKA-31 (naphtho[1,2-d]thiazol-2-ylamine), a compound which was developed in our laboratory using the neuroprotectant riluzole as a structural template (Sankaranarayanan et al., 2009). Although SKA-31 is more potent than EBIO and activates KCa3.1 with an EC50 of 260 nM and KCa2 channels with EC50 values of 1.9 and 2.9 μM, it also still suffers from a lack of selectivity. Through further derivatization of SKA-31, we more recently identified SKA-121 (5-methylnaphtho[2,1-d]oxazol-2-amine), the first selective KCa3.1 activator (Coleman et al., 2014). SKA-121 displays 40-fold selectivity for KCa3.1 (EC50 = 110 nM) over KCa2.3 (EC50 = 4.4 μM) and slightly greater selectivity over KCa2.2 and KCa2.1. Interestingly, this gain in selectivity for KCa3.1 over KCa2.3 was achieved by a very minor structural change, addition of a methyl group in five-position of the napthothiazole or the related napthooxazole ring system. We were very intrigued to observe this “magic methyl effect,” meaning that the installation of a single methyl group in the “right” location on a chemical scaffold can result in a significant activity shift (Leung et al., 2012), and therefore, we wanted to identify the binding site of the napthothiazole/oxazole-type KCa activators and, if possible, understand this selectivity at the molecular level.

Using the recently determined X-ray crystal structures of the KCa2.2 CaM-BD in complex with CaM and containing the KCa activators EBIO (Zhang et al., 2012) or NS309 (6,7-dichloro-1H-indole-2,3-dione 3-oxime) (Zhang et al., 2013), we used a combination of RosettaLigand molecular modeling (https://www.rosettacommons.org/) and site-directed mutagenesis to generate homology models of the KCa3.1 and KCa2.3 CaM-BD/CaM complex containing SKA-121, SKA-111 (5-methylnaphtho[1,2-d]thiazol-2-amine), SKA-31, EBIO, and NS309 in the interface. KCa3.1 selectivity seems to be due to the presence of R362, a residue which is at the center of an extensive hydrogen bond network stabilizing SKA-121 in KCa3.1. If R362 in KCa3.1 is replaced by a serine like in the KCa3.1-R362S mutant or in KCa2.3, the hydrogen bond network is not present.

Materials and Methods

Chemicals and Reagents.

SKA-121 and SKA-31 were synthesized in the Wulff laboratory as previously described and characterized by 1H and 13C NMR for identity and by ultra-performance liquid chromatography/mass spectrometry for purity (Sankaranarayanan et al., 2009; Coleman et al., 2014). EBIO and NS309 were purchased from Sigma-Aldrich (St. Louis, MO). Standard laboratory chemicals were purchased from commercial vendors and were of the purest available grade. KCa activator stocks (1 or 10 mM) were prepared in dry dimethylsulfoxide, and drug dilutions for all electrophysiology experiments were always freshly prepared during the experiment to avoid any potential precipitation of compounds. The final dimethylsulfoxide concentration was typically around 0.1% and never exceeded 1%.

Molecular Biology.

The cloning of human KCa3.1 (#AF033021) and KCa2.3 (#AJ251016) was reported in the late 1990s (Logsdon et al., 1997; Chandy et al., 1998). Both channel genes were subcloned in-frame downstream to green fluorescent protein in the pEGFP-C1 expression vector (CLONTECH, Mountain View, CA) (Wulff et al., 2001). All clones were verified by sequencing. Mutations were introduced using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) and were verified by fluorescence sequencing. For amino acid numbering for KCa3.1, we used the following gene: Homo sapiens potassium channel, calcium-activated intermediate-/small-conductance subfamily N alpha, member 4 (KCNN4; National Center for Biotechnology Information reference sequence: NP_002241.1). For amino acid numbering for KCa2.3, we used the following gene: Homo sapiens potassium channel, calcium-activated intermediate-/small-conductance subfamily N alpha, member 3 (KCNN3), transcript variant 1, mRNA (National Center for Biotechnology Information reference sequence: NM_002249.5). This corresponds to the Q14 variant of KCa2.3.

Electrophysiology.

All experiments were performed in either the inside-out or the whole-cell configuration of the patch-clamp technique on transiently transfected COS-7 cells. COS-7 cells were purchased from American Type Culture Collection (Manassas, VA) and cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum. Wild-type (WT) and mutant KCa3.1 and KCa2.3 channel constructs were transfected using FuGENE 6 transfection reagent (Promega, Madison, WI) in Opti-MEM reduced-serum medium (Life Technologies, Benicia, CA). Cells were cultured in six-well plates for 48 hours and then detached by TrypLE Express (Gibco, Grand Island, NY) and plated on coverslips for 30 minutes to 1 hour for whole-cell recordings. For inside-out recordings, cells were plated 2–3 hours before the experiments to attach them more firmly. Coverslips were placed in a 15-μl recording chamber mounted on an inverted microscope (Olympus XI-70 equipped with fluorescence burner and filters; Olympus, Tokyo, Japan), and only clearly green fluorescent cells were patch-clamped. To reduce contaminating currents from native chloride channels in COS-7 cells, solutions contained aspartate instead of chloride. For whole-cell experiments, the extracellular solution contained 160 mM Na+-aspartate, 4.5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, and 10 mM HEPES (pH 7.4, 300 mOsm). Solutions on the intracellular side contained 160 mM K+-aspartate, 10 mM HEPES, 10 mM EGTA, 2.31 mM MgCl2, and 5.96 or 7.2 mM CaCl2 for a calculated free Ca2+ concentration of 250 or 500 nM (pH 7.2, 290 mOsm). For inside-out experiments in symmetrical K+, the extracellular solution contained 160 mM K+-aspartate, 10 mM HEPES, 1 mM MgCl2, and 2 mM CaCl2 (pH 7.4, 300 mOsm). Intracellular solutions contained 160 mM K+-aspartate, 10 mM HEPES, 10 mM EGTA, 2.31 mM MgCl2, and varying amounts of CaCl2 for calculated free Ca2+ concentrations of 0.05, 0.1, 0.25, 0.5, 1, 10, and 30 μM (pH 7.2, 300 mOsm). Free Ca2+ concentrations were calculated using the July 3, 2009 online version of MaxChelator (http://maxchelator.stanford.edu/webmaxc/webmaxcS.htm) assuming a temperature of 25°C, a pH of 7.2, and an ionic strength of 160 mM. Patch pipettes were pulled from soda lime glass (micro-hematocrit tubes; Kimble Chase, Rochester, NY) and had resistances of 1.5–3 MΩ when submerged. Experiments were controlled with a HEKA EPC-10 amplifier and Pulse software (HEKA, Lambrecht/Pfalz, Germany). In whole-cell experiments, cells were clamped to a holding potential of −80 mV, and KCa currents were elicited by 200-ms voltage ramps from −120 to +40 mV applied every 10 seconds. In inside-out experiments, cells were clamped to a holding potential of −80 mV, and KCa currents were elicited by 200-ms voltage ramps from −80 to +80 mV applied every 5 seconds. Data analysis, fitting, and plotting were performed with IGOR-Pro (Wavemetrics, Lake Oswego, OR) or Origin 9 (OriginLab Corporation, Northampton, MA).

Calcium Sensitivity Testing.

To assess whether specific mutations affected Ca2+ sensitivity, mutant channels were first patched with an intracellular solution containing 3 μM of free Ca2+. If the mutant did not display currents with amplitudes in the nA range under these conditions, the mutant was deemed to have altered Ca2+ sensitivity and was excluded from subsequent experiments investigating sensitivity to SKA-121.

Homology Modeling.

We generated preliminary models of the KCa3.1 channel C-terminal CaM-binding domain in complex with CaM using Rosetta computational modeling software (Rohl et al., 2004; Yarov-Yarovoy et al., 2006; Wang et al., 2007; Mandell et al., 2009) and an X-ray structure of the KCa2.2 channel CaM-binding domain in complex with CaM and NS309 (Zhang et al., 2013) [Protein Data Bank (pdb) identifier: 4J9Z). Due to significant sequence differences between KCa3.1 and KCa2 channels in the loop between the two helices forming the CaM-binding domain, we predicted the structure of this region de novo using Rosetta cyclic coordinate descent and kinematic closure loop modeling methods developed to model loop structures with subangstrom accuracy (Wang et al., 2007; Mandell et al., 2009). After performing several rounds of cyclic coordinate descent and kinematic closure loop modeling with at least 10,000 models generated during each round, models were ranked based on total Rosetta energy (Rohl et al., 2004). Using a root mean square deviation threshold that placed 1%–2% of all models in at least one of the largest clusters, 10% of the lowest energy models were clustered (Bonneau et al., 2002). Models representing centers of the top 20 clusters (early rounds) and/or the best 10 models by total energy (later rounds) were used as input for the next round of loop modeling. Potential differences in backbone and side-chain conformations of KCa3.1 compared with the KCa2.2 structure template were explored using Rosetta’s full-atom relax protocol (Rohl et al., 2004; Barth et al., 2007). Selection of the best KCa3.1 channel CaM-binding domain in complex with CaM models was guided by clustering of the lowest energy models to generate the most frequently sampled ensembles of models (Bonneau et al., 2002). A Rosetta model of the KCa2.3 channel C-terminal CaM-binding domain in complex with CaM was generated using the same procedure.

Ligand Docking.

Ligand docking was performed using the Rosetta-Ligand method (RosettaLigand application from Rosetta program suite, version 3.7), which progresses in three stages from low-resolution conformational sampling and scoring to full-atom optimization using the all-atom energy function (Meiler and Baker, 2006; Davis and Baker, 2009; Davis et al., 2009). In the first, low-resolution stage, the ligand was placed randomly within the binding site, and its “center of mass” was constrained to move within a 10-Å diameter sphere. Ligand conformers were generated using Open Eye OMEGA software, version 2.5.1.4 (OpenEye Scientific Software, Santa Fe, NM) (Hawkins et al., 2010; Hawkins and Nicholls, 2012; http://www.eyesopen.com) and were then randomly rotated as a rigid body and scored for shape compatibility with the protein. The best-scoring models were filtered by root mean square deviation to eliminate near duplicates, and one of the remaining models was selected at random to continue to the next stage. The second, high-resolution stage used the Monte Carlo minimization protocol in which the ligand position and orientation were randomly perturbed by a small deviation (0.1 Å and 3°); receptor side chains were repacked using a rotamer library; the ligand position, orientation, and torsions and protein side-chain torsions were simultaneously optimized using quasi-Newton minimization; and the end result was accepted or rejected based on the Metropolis criterion. Scoring used the full-atom Rosetta energy function with softened van der Waals repulsion. The side-chain rotamers were searched simultaneously during “full repack” cycles and one at a time in the “rotamer trials” cycles. The full repack made ∼106 random rotamer substitutions at random positions and accepted or rejected each based on the Metropolis criterion. Rotamer trials chose the single-best rotamer at a random position in the context of the current state of the rest of the system, with the positions visited once each in random order. The ligand was treated as a single residue and its input conformers served as rotamers during this stage. During the energy minimization step, the finely sampled rotamer library and soft-repulsive energy function allow access to off-rotamer conformations. The third and final stage was a more stringent gradient-based minimization of the ligand position, orientation, and torsions and the receptor torsions for both side chains and backbone. Scoring used the same Rosetta energy function but with a hard-repulsive van der Waals potential, which created a more rugged energy landscape that was better at discriminating native from non-native binding modes. A total of 10,000 models were generated for each docking trial using a University of California, Davis cluster system. To identify the best model, the models were first screened by total energy score (Rosetta energy term: score), and the top 1000 models with the lowest total energy score were selected. They were further scored by the binding energy between ligand and channel. Binding energy (interface_delta_X) was calculated as the difference in total energy between the complex with the ligand bound and the complex with the ligand separated from the binding site. The top 10 models with the lowest binding energy were identified as the candidates that exhibited good structural convergence. All molecular graphics of ligand and KCa channel C-terminal CaM-binding domain in complex with CaM were rendered using the UCSF Chimera software (Resource for Biocomputing, Visualization, and Informatics, San Francisco, CA) (Pettersen et al., 2004).

Protein Data Bank format files of the Rosetta models of the KCa3.1 and KCa2.3 channel CaM-BD/CaM with and without SKA-121 in the interface are provided in the Supplemental Material; pdb files of all other models are available upon request.

Results

The Positive Gating Modulator SKA-121 Increases the Open-Probability of KCa3.1 and KCa2.3.

Positive gating modulators of KCa channels such as EBIO and NS309 increase channel open probability (PO) at a given intracellular Ca2+ concentration, thus “potentiating” the current and resulting in a leftward shift of the apparent Ca2+ concentration-response curve (Pedarzani et al., 2001; Li et al., 2009; Zhang et al., 2013). We previously showed that SKA-121 has similar potentiating effects on KCa3.1 by increasing PO in a calcium-dependent manner and displays ∼40-fold selectivity for KCa3.1 over KCa2.3 (Coleman et al., 2014). Accordingly, a concentration of 1 μM SKA-121 maximally activates KCa3.1 while only minimally activating KCa2.3 (Fig. 1, A and B). What was not investigated previously was whether SKA-121 also increases the open channel probability and shifts the Ca2+ concentration-response curve of KCa2 channels. We therefore performed inside-out experiments on COS-7 cells stably expressing human KCa2.3, varied the intracellular [Ca2+]i concentration, and found free calcium EC50 values of 1.19 μM in the absence of SKA-121 and 0.41 μM in the presence of 20 μM SKA-121 (Fig. 1D). Interestingly, in addition to a clear left shift of the Ca2+ concentration activation response curve of KCa2.3, there also was an increase in maximal effect even at free Ca2+ concentrations of 10 and 30 μM. This was somewhat surprising, because unlike KCa3.1 channels, which have a relatively low Ca2+-dependent Po(max) (Gerlach et al., 2001; Jones et al., 2007), KCa2 channels are supposedly fully open at saturating [Ca2+]i concentrations. The Ca2+ concentration activation response curve for KCa3.1 is shown for comparison (Fig. 1C). The free calcium EC50 for KCa3.1 was found to be 0.59 μM, and shifted to 0.42 μM in the presence of 500 nM SKA-121 with a simultaneous 2-fold increase in the maximally achievable current (Fig. 1C). These data suggest that SKA-121, although more potent on KCa3.1 and inducing a larger increase in PO under comparable recording conditions, acts as a positive gating modulator that “potentiates” current in a Ca2+-dependent manner on both KCa3.1 and KCa2.3.

Fig. 1.
Fig. 1.

SKA-121 displays 40-fold selectivity for KCa3.1 over KCa2.3 but is a positive gating modulator for both channels. (A) Whole-cell KCa3.1 current with an intracellular free calcium concentration of 250 nM in the presence and absence of 1 μM SKA-121. (B) Whole-cell KCa2.3 current with an intracellular free calcium concentration of 500 nM in the presence and absence of 1 μM SKA-121. Both traces in (A and B) were elicited by voltage ramps from −120 to + 40 mV applied every 10 seconds. Inside-out KCa channel experiments: (C) Calcium concentration-response curves for KCa3.1 in the presence [EC50 = 424 nM, 95% confidence interval (CI): 399–454 nM, nH = 3.1] and absence (EC50 = 590 nM, 95% CI: 556–625 nM, nH = 2.9) of 500 nM SKA-121. (D) Calcium concentration response curves for KCa2.3 in the presence (EC50 = 406 nM, 95% CI: 318–518 nM, nH = 2.2) and absence (EC50 = 1.189 μM, 95% CI: 0.999–1.904 nM, nH = 2.7) of 20 μM SKA-121. Data points are the mean ± S.D.; n = 3–5 independent cells/data point. An extra sum-of-squares F test (GraphPad Prism5; GraphPad Software, La Jolla, CA) to compare the curves rendered P values <0.0001 for the comparison of both the KCa3.1 and the KCa2.3 calcium sensitivity in the presence and absence of SKA-121.

The CaM-BD Is the Functional Binding Site for KCa Channel Activators.

Based on their similar mechanism of action and the overall structural similarity between napthothiazole/oxazole-type KCa activators such as SKA-31 and SKA-121 and benzimidazole-type activators such as EBIO and NS309, we suspected that SKA-121 binds at the interface between CaM and the CaM-BD. Zhang et al. (2012) soaked EBIO into the protein crystal of CaM bound to the CaM-BD of KCa2.2. In a subsequent study, the same group obtained a cocrystal of the CaM-BD/CaM with NS309 (Zhang et al., 2013). Both molecules reside in a pocket formed at the interface between CaM-BD/CaM, and the more potent NS309 seems to facilitate channel gating by affecting the conformation of an intrinsically disordered stretch of nine amino acids, which connects transmembrane segment S6 to the CaM-BD (Zhang et al., 2013).

Since the sequences of KCa2.2 and KCa2.3 in the CaM-BD are similar to each other but differ from KCa3.1 (Fig. 2A), we performed site-directed mutagenesis to determine whether the CaM-BD/CaM interface identified in the Zhang work (Zhang et al., 2012, 2013) also represented the binding site of the napthothiazole/oxazole-type KCa activators such as SKA-31 and SKA-121. As a first step, we created the KCa2.3 mutation A625I and the double mutation A625V/L628M because the corresponding KCa2.2 mutations (A477I and A477V/L480M) had been shown by Zhang et al. (2012) to modify EBIO activity. Whereas the double mutant (which makes KCa2.2 more KCa3.1-“like”) had been reported to be more EBIO sensitive, the single mutant had made KCa2.2 less sensitive to EBIO. For KCa2.3, the corresponding mutants both decreased EBIO (not shown) and SKA-31 sensitivity (Fig. 2B), suggesting that the CaM-BD/CaM interface is also the binding site for the napthothiazole/oxazole-type KCa activators but that there are subtle differences between the individual KCa channels and/or the exact orientation of the different activators in the interface pocket.

Fig. 2.
Fig. 2.

(A) Sequence alignment of the calcium-binding domain of KCa2.2, KCa2.3, and KCa3.1. Zhang et al. (2012) mutated the highlighted residues in KCa2.2, A to V and L to M, to the corresponding KCa3.1 residues and demonstrated increased EBIO sensitivity. (B) Responses of the KCa2.3–CaM-BD mutants A625V/L628M and A625I to SKA-31. Whereas SKA-31 at a concentration of 1 μM does not elicit an increase in current for all three channels, 5 and 10 μM SKA-31 produce strong current increases in the WT-KCa2.3 channel that are significantly reduced in both the A625V/L628M and A625I-KCa2.3 mutants. (C) Summary of the initial KCa3.1 whole-cell experiments in response to SKA-121. Position V365 does not appear to be involved in SKA-121 binding since current increases in response to 500 nM SKA-121 compared with WT-KCa3.1 are not statistically different for both the V365A and V365F mutants. The M368L mutant, in contrast, displays an increased current response to 500 nM SKA-121 (P = 4.076E−4), suggesting that M368 is involved in SKA-121 binding. All values are the mean ± S.D.; n = 5–8 cells per mutant channel (two-sample t test assuming equal variance). *P < 0.05; **P < 0.01; ***P < 0.001.

We next switched from the prototype compound SKA-31 to the KCa3.1 selective compound SKA-121 and tested how mutations of the corresponding positions, V365 and M368 in KCa3.1, would affect SKA-121 potency. We created four mutants: V365A, V365F, M368A, and M368L. The mutant V365F was generated to incorporate a bulky amino acid into the binding pocket to potentially disturb SKA-121 binding, and M368A was created to remove potential SKA-121 interaction points. All four mutations exhibited normal Ca2+ sensitivity, and the responses to 500 nM SKA-121 are summarized in Fig. 2C. We had expected that the V365A and V365F mutants would be less sensitive to SKA-121 than the WT channel. However, V365A, V365F, and M368A were not statistically different in their responses to 500 nM SKA-121 [29.20 ± 18.70, 29.07 ± 9.45, and 31.24 ± 16.05 fold increase (FI) in current, respectively] compared with WT (25.89 ± 5.93 FI). In contrast, 500 nM SKA-121 activated the M368L mutant (66.55 ± 9.90 FI) more effectively than the WT channel. These results suggest that SKA-121 probably sits somewhat differently in the KCa3.1 binding pocket than EBIO sits in the KCa2.2 binding pocket and, therefore, does not make contact with V365, which corresponds to A625 in KCa2.3 and A477 in KCa2.2. There instead seems to be an interaction with M368 (see Fig. 7). However, taken together with the aforementioned KCa2.3 data, the results strongly suggest that SKA-121 and SKA-31 bind at the CaM-BD/CaM interface similar to EBIO.

We next generated Rosetta homology models of the CaM-BD/CaM of KCa2.3 and KCa3.1 based on the KCa2.2 CaM-BD/CaM crystal structure (pdb identifier: 4J9Z) (Zhang et al., 2013) and then used RosettaLigand computational modeling software to dock SKA-121 into the binding pocket formed at the CaM-BD/CaM interface, as described in Materials and Methods. The KCa3.1 model with the channel (green) and calmodulin (pink) in a space-filling representation with SKA-121 docked at the CaM-BD/CaM interface is shown in Fig. 3A (pdb file provided in Supplemental Materials). We probed the model by mutating residues that were facing inward into the putative binding pocket along a stretch of 11 amino acids (362–372 in KCa3.1) in the CaM-BD of both KCa2.3 and KCa3.1 (Fig. 3, B–D; pdb files provided in Supplemental Materials) and determined SKA-121 activity after first studying Ca2+ sensitivity of the mutants and excluding mutants not responding with nA currents to free intracellular Ca2+ concentrations of 3 μM. Overall, most mutations (Figs. 3, C and D and 2C for KCa3.1) did not significantly change the response of KCa2.3 to 20 μM SKA-121 or of KCa3.1 to 500 nM SKA-121. The KCa2.3 mutants D623E and V629A, as well as the KCa3.1 mutants M368F and V365A/S367T/M368L were not functional, even at saturating calcium concentrations, and therefore could not be studied. As suggested by the models, substituting a conserved Ser facing into the binding pocket (S632 in KCa2.3 and S372 in KCa3.1) with an Arg made both KCa2.3 and KCa3.1 insensitive to SKA-121 (Fig. 3, C and D). The only other interesting mutation was KCa3.1 R362S, which was significantly less sensitive to SKA-121 than WT KCa3.1 (Fig. 3D).

Fig. 3.
Fig. 3.

(A) Space-fill rendering of the Rosetta model of SKA-121 docked into the interface between the KCa3.1 CaM-BD (light green) and CaM (pink). (B) Sequence alignment of the CaM-BD of KCa3.1 and KCa2.3. (C) Effect of KCa2.3 mutations on the current activating effect of 20 μM SKA-121. Shown are fold increases in comparison with control currents elicited by 500 nM free internal Ca2+. (D) Effect of KCa3.1 mutations on the current activating effect of 500 nM SKA-121. Shown are fold increases in comparison with control currents elicited by 250 nM free internal Ca2+. Shown are the mean ± S.D.; n = 5–8 cells per mutant (two-sample t test assuming equal variance). *P < 0.05; ***P < 0.001.

Mutational Confirmation of the KCa Activator Binding Pocket in KCa3.1.

Since the most dramatic effect was observed when mutating S372 in KCa3.1 to a charged and bulky Arg, we hypothesized that this mutation would prevent activators from binding to the interface pocket and should render KCa3.1 generally insensitive to both benzimidazole-type activators such as EBIO and NS309 as well as other napthothiazole/oxazole-type KCa activators such as SKA-31 and SKA-111. Interestingly, the S372R mutation practically abolished sensitivity to SKA-31 (1 μM, Fig. 4A), SKA-111 (1 μM, Fig. 4B), and EBIO (100 μM, Fig. 4C) in comparison with the WT KCa3.1 channel, but did not significantly (P = 0.095) reduce the sensitivity of NS309 (1 μM, Fig. 4D). Rosetta models of the four compounds docked and energy minimized in CaM-BD/CaM of KCa3.1 were found to be in agreement with these observations. The lowest energy-binding poses of SKA-111 (EC50 = 110 nM) and SKA-31 (EC50 = 260 nM) show two hydrogen bonds between the two-position amino group and M51 and E54 in calmodulin, whereas E54 itself is part of an extensive hydrogen bond network including R362, E295, and N300 in KCa3.1 (Fig. 4, A and B). The less potent EBIO only forms one hydrogen bond through its one available hydrogen bond donor on the imidazole nitrogen with the backbone carbonyl oxygen on M51 (Fig. 4C). NS309, which is the most potent KCa3.1 activator of all four compounds (EC50 = ∼∼30 nM) forms the same hydrogen bond with the backbone carbonyl oxygen on M51 through its indole NH but also extends its three-position oxime group toward R362 and accepts a hydrogen bond from one of the NH2 groups of the guanidine moiety of R362 (Fig. 4D).

Fig. 4.
Fig. 4.

The KCa3.1-S372R mutation significantly reduces the potency of SKA-31 (P = 0.024) (A), SKA-111 (P = 0.004) (B), and EBIO (P = 0.0002) (C), but not NS309 (P = 0.095) (D). Shown are fold increases in current (mean ± S.D.) compared with WT control at 250 nM free internal [Ca2+] from three to six independent cells per condition (two-sample t test assuming equal variance). *P < 0.05; **P < 0.01; ***P < 0.001.

R362 Is a Key Residue for the Greater KCa3.1 Potency of SKA-121 and NS309.

After thus confirming the general location of the napthothiazole/oxazole binding pocket, we further investigated the KCa3.1 R362S mutation since this was the only one of our mutations that had selectively reduced SKA-121 potency on KCa3.1 (Fig. 3D). To understand this loss of potency, we modeled the KCa3.1-R362S mutant and closely compared the lowest energy conformations of SKA-121 between WT-KCa3.1 (Fig. 5A, pdb provided in Supplemental Material), KCa3.1-R362S (Fig. 5B), and WT-KCa2.3 (Fig. 5C; pdb provided in Supplemental Data). In the WT-KCa3.1 model, the amino group of SKA-121 forms hydrogen bonds with the backbone carbonyl oxygen on M51 and the side-chain oxygen on E54 in calmodulin, whereas the E54 side-chain oxygen itself is part of an extensive hydrogen bond network including R362, E295, and N300 in the KCa3.1 channel (Fig. 5A). If R362 in KCa3.1 is replaced by Ser, as in the KCa3.1-R362S mutant (Fig. 5B), the two hydrogen bonds to M51 and E54 are still present in the majority of low energy-binding poses, but the hydrogen bond network that involved R362 in the WT-KCa3.1 (Fig. 5A) is absent. Similarly, the KCa2.3 model is devoid of this hydrogen bond network (Fig. 5C). Full concentration-response curves (Fig. 5D) revealed that the KCa3.1-R362S mutant is ∼7-fold less sensitive to SKA-121 than the WT-KCa3.1 channel, suggesting that R362 is one of the key residues responsible for the KCa3.1 selectivity of SKA-121, presumably because of its ability to stabilize E54 in CaM through a hydrogen bond network. In keeping with this idea, the KCa3.1-R362S mutant also significantly reduced the potency of SKA-31 and SKA-111 (Supplemental Fig. 1). Conversely, the reverse mutation in KCa2.3 (KCa2.3-S622R) only slightly, but not significantly, increased the sensitivity of KCa2.3 to SKA-121 (EC50 = 8.75 vs. 10.25 μM), suggesting that the presence of an Arg residue alone is not sufficient to generate a high-affinity binding site for small-molecule KCa activators in the KCa2.3–CaM-BD/CaM interface. Figure 6 shows an overlay of the lowest energy-binding poses of NS309 (orange) and SKA-121 (dark green), illustrating that NS309 sits “more forward” in the CaM-BD/CaM interface of KCa3.1 than SKA-121 and forms a direct hydrogen bond with R362.

Fig. 5.
Fig. 5.

Rosetta models of the lowest energy-binding poses of SKA-121 in the interface between CaM (pink) and CaM-BD (light green) of WT-KCa3.1 (A), the KCa3.1-R362S mutant (B), and WT-KCa2.3 (C). (D) Concentration-response curves for SKA-121 induced current activation: WT-KCa3.1 [EC50 = 128 nM, 95% confidence interval (CI): 84–194 nM, nH = 1.3), KCa3.1-R362S (EC50 = 851 nM, 95% CI: 0.662–1.288 μM, nH = 1.3), KCa2.3-A625V (EC50 = 4.73 μM, 95% CI: 2.84–5.72 μM, nH = 2.8), KCa2.3-S622R (EC50 = 8.45 μM, 95% CI: 7.42–9.55 μM, nH = 2.8), WT-KCa2.3 (EC50 = 10.25 μM, 95% CI: 6.16–33.59 μM, nH = 2.7). Data points are expressed as the mean ± S.D.; n = 3–5 independent cells/data point. An extra sum-of-squares F test (GraphPad Prism5) to compare the curves rendered a P value <0.0001 for the comparison between WT-KCa3.1 and the KCa3.1-R362S mutant. Comparisons of the KCa2.3 mutants to WT-KCa2.3 showed that the SKA-121 EC50 was significantly different for the KCa2.3-A625V mutant (P = 0.0091) but not the KCa2.3-S622R (P = 0.6636) mutant.

Fig. 6.
Fig. 6.

Overlay of the lowest energy-binding poses of NS309 (orange) and SKA-121 (dark green) in the interface between the KCa3.1–CaM-BD (light green) and CaM (pink). Although R362 in KCa3.1 does not directly interact with SKA-121, it stabilizes E54 in CaM through three hydrogen bonds. NS309, in contrast, “sits more forward” and is stabilized through a direct hydrogen bond with R362 and M51. The inset on the right shows an overlay of NS309 and SKA-121 with all channel residues removed for clarity.

The KCa3.1 Selectivity of SKA-121 Is Due to a Better Overall Hydrophobic Shape Complementarity.

Although the direct or indirect hydrogen bonding to R362 in KCa3.1 could thus be postulated to explain the greater KCa3.1 potency of both napthothiazole/oxazole-type KCa activators and NS309, this contact could still not explain the greater KCa3.1 selectivity of SKA-121 and SKA-111 when compared with their template SKA-31, since all three compounds, SKA-31 (Fig. 4A), SKA-111 (Fig. 4B), and SKA-121 (Fig. 5A), form two hydrogen bonds with M51 and E54 in calmodulin with a “background” stabilization of E54 from a hydrogen bond to R362. Therefore, to understand why the introduction of a methyl group in the five-position of the napthothiazole/oxazole ring system resulted in a roughly 2.5-fold gain in potency on KCa3.1 and a roughly 3-fold (SKA-121) drop in potency on KCa2.3 compared with SKA-31 (EC50 = 2.9 μM for KCa2.3), we more closely scrutinized the hydrophobic protein ligand interactions in the CaM-BD/CaM interface for SKA-31 and SKA-121 in our Rosetta models (Fig. 7). The most prominent hydrophobic interactions formed by SKA-31 in the KCa3.1–CaM-BD/CaM interface involve the sulfur atom in its thiazole ring, which interacts with both M71 in CaM and V365 in KCa3.1, and the outer ring of the napthol system, which makes hydrophobic contacts with F19 and I63 in CaM and M368 in KCa3.1 (Fig. 7A). To experimentally confirm the interaction between the thiazole sulfur and V365, we tested SKA-31 on both the V365A and the V365F mutants and found that V365A was roughly 2-fold more sensitive, whereas the V365F mutation, which had not affected SKA-121 activity (Fig. 3), completely abolished SKA-31 but not NS309 activity (data not shown). The remaining van der Waals interactions between the thiazole amino group of SKA-31 and M51 and E54 further strengthen the aforementioned hydrogen bonds to these two CaM residues (Fig. 7A). SKA-121, instead of contacting V365 and M71, only contacts M71 through its oxazole nitrogen but “picks up” an additional contact with M51 through its oxazole oxygen and three hydrophobic interactions between its five-position methyl group and S372, M368, and M72 (Fig. 7C). A wire-mesh rendering of the surface area of SKA-121 in the KCa3.1–CaM-BD/CaM interface pocket (Fig. 7B) illustrates a nearly perfect shape complementarity between the –CH3 group of SKA-121 and the pocket formed by S372, M368, and M72 in the channel/CaM interface. SKA-31, in contrast, does not fill this pocket as tightly, which could account for its ∼2.5-fold lower potency in activating KCa3.1 (Fig. 7D).

Fig. 7.
Fig. 7.

Hydrophobic interactions (dark purple) in the lowest energy-binding poses of SKA-31 (A) and SKA-121 (C) in the KCa3.1–CaM-BD/CaM interface. KCa3.1 (light green) or CaM (pink) residues within a 5-Å radius of the ligand undergoing hydrophobic interactions are shown explicitly in a stick rendering. Residues not forming interactions are hidden for clarity. The hydrogen bonds to M51 and E54 in CaM are also not shown for clarity. Wire-mesh rendering of the surface area of SKA-31 (B) and SKA-121 (D) in the KCa3.1–CaM-BD/CaM interface pocket viewed from the same angle as in (A and B). (E and F) Hydrophobic interactions (dark purple) in the lowest energy-binding poses of SKA-31 (E) and SKA-121 (F) in the KCa2.3–CaM-BD/CaM interface.

A detailed look at the hydrophobic interactions between SKA-31 and SKA-121 and the KCa2.3–CaM-BD/CaM interface pocket reveals that both compounds make fewer hydrophobic contacts (six for SKA-31 and five for SKA-121) than in the KCa3.1 interface.

Discussion

KCa channels have a well developed pharmacology in comparison with many other ion channels. KCa2 channels are blocked by the venom peptides apamin, scyllatoxin and tamapin and the bisquaternary small molecules UCL-1684 (6,10-diaza-3(1,3)8(1,4)-dibenzena-1,5(1,4)-diquinolinacyclodecaphane) and UCL-1848 (6,10-diaza-3(1,3)8(1,4)-dibenzena-1,5(1,4)-diquinolinacyclodecaphane) (Wulff et al., 2007). KCa3.1 channels, in contrast, are inhibited by the scorpion toxins maurotoxin (Castle et al., 2003) and charybdotoxin (Rauer et al., 2000), the triarylmethanes TRAM-34 (1-[(2-chlorophenyl)diphenylmethyl]-1H-pyrazole) (Wulff et al., 2000) and senicapoc (Stocker et al., 2003), as well as the benzothiazinone NS6180 (4-[[3-(trifluoromethyl)phenyl]methyl]-2H-1,4-benzothiazin-3(4H)-one) (Strøbæk et al., 2013). In addition to these “simple” inhibitors, which bind either in the outer pore similar to the toxins or in the inner pore similar to the triarylmethanes and NS6180 (Nguyen et al., 2017), KCa2/3 channels also have positive and negative gating modulators, which apparently render the channels more or less Ca2+-sensitive by left shifting or right shifting the Ca2+-response curve (Christophersen and Wulff, 2015). Whereas negative gating modulators such as NS8593 ((R)-N-(benzimidazol-2-yl)-1,2,3,4-tetrahydro-1-naphtylamine) and (−)CM-TMPF interact with positions deep within the inner pore vestibule (Jenkins et al., 2011; Hougaard et al., 2012), where the gate of these channels is located (Bruening-Wright et al., 2002, 2007), positive gating modulators have long been suspected to act through a site located in the C-terminal region, close to or at the CaM-BD based on work from Pedarzani et al. (2001), who showed that swapping of the C terminus could transfer the higher EBIO sensitivity from KCa3.1 to KCa2.2. This prediction was more recently confirmed when Zhang et al. crystallized the KCa2.2 CaM-BD/CaM complex first with EBIO (Zhang et al., 2012) and then with NS309 (Zhang et al., 2013) in the interface of the CaM-BD and the CaM N-lobe. Since both EBIO and NS309 are unsuitable for in vivo experiments due to their low potency or metabolic instability, our group developed a series of KCa2/3 activators based on the more “drug-like” napthothiazole/oxazole scaffold (Sankaranarayanan et al., 2009; Coleman et al., 2014). Despite its much better pharmacokinetic properties, the first-generation compound, SKA-31, only displayed a 7-fold selectivity for KCa3.1 over KCa2 channels, which is very similar to EBIO. In contrast, two of our second-generation compounds, SKA-121 and SKA-111, displayed greatly improved selectivity for KCa3.1 over KCa2.3, which was caused by the introduction of a methyl group in the five-position of the flat napthothiazole/oxazole ring system (Coleman et al., 2014).

Here, we set out to understand this “magic methyl” effect (Leung et al., 2012) at the molecular level. Based on the overall structural similarity between SKA-31 and SKA-121 and EBIO and NS309, we strongly suspected that napthothiazole/oxazole-type KCa activators also bind at the CaM-BD/CaM interface. However, since we had previously been totally mistaken in the assumption that the negative gating modulator NS8593, which turned out to have a binding site in the inner pore despite “looking” similar to EBIO and NS309 (Jenkins et al., 2011; Hougaard et al., 2012), binds at this interface, we first used site-directed mutagenesis to confirm that this assumption was indeed valid. After localizing the binding site of SKA-31 and SKA-121 to the KCa channel CaM-BD/CaM interface, we generated homology models of the KCa3.1 and KCa2.3 CaM-BD/CaM complexes based on the crystal structure of the KCa2.2 CaM-BD/CaM complex (Zhang et al., 2013) with the RosettaLigand computational modeling software and then used these models to further guide experimentation to explore the orientation of SKA-121 and other KCa activators in their binding pocket. Taken together, our modeling and mutagenesis data suggest that all KCa activators (EBIO, NS309, SKA-31, and SKA-121) hydrogen bond with CaM-M51 in both KCa3.1 and KCa2.3 channels. This hydrogen bond is very likely also present in the KCa2.2 CaM-BD/CaM crystal structure (Zhang et al., 2013) (pdb identifier: 4J9Z), where the hydrogen on the imidazole nitrogen of NS309 is in perfect hydrogen bonding distance to the oxygen of the backbone carbonyl group of CaM-M51. The more potent napthothiazole/oxazole-type KCa activators SKA-31, SKA-111, and SKA-121 (Fig. 4) all form an additional second hydrogen bond with E54 and thus “anchor” themselves with their two-position amino group on two carbonyl groups of the calmodulin N-lobe. This second hydrogen bond is present in all of our KCa3.1 and KCa2.3 docking models of KCa activators with an amino group (for example, see SKA-121 in KCa2.3 in Fig. 5C). What differentiates the most frequently sampled lowest energy conformations of the KCa3.1 models with napthothiazole/oxazole-type KCa activators in the CaM-BD/CaM interface is the presence of an extensive hydrogen bond network with KCa3.1-R362 at its center that seems to stabilize CaM-E54 in an optimal position to hydrogen bond with the amino group of the docked ligand (Figs. 4 and 5). We therefore postulate that the presence of R362 in KCa3.1 is responsible for the generally observed 5- to 10-fold KCa3.1 selectivity of all KCa activators. NS309, the most potent KCa3.1 activator, directly hydrogen bonds with R362.

Having found a reasonable structural explanation in the presence of R362 for why KCa3.1 is more sensitive to SKA-31, SKA-121, SKA-111, and NS309 than KCa2.3, we next focused our attention on the CH3 group in the five-position of the napthothiazole/oxazole ring system of SKA-121 and SKA-111. Interestingly, on closer inspection, the effect of this group is not exactly what is classically termed the “magic methyl” effect (Leung et al., 2012), where the addition of a single methyl group in the “right” location on a scaffold results in a dramatic increase in activity, but rather a relatively modest 2.5-fold increase in potency on KCa3.1 combined with a moderate three- to five-fold loss in potency on KCa2.3, which overall results in a 40- to 100-fold increase in KCa3.1 selectivity compared with SKA-31. Given that these changes, although quite effective when combined, do not suggest major steric hindrance from stabilization of, for example, a large rotatable ring (Barreiro et al., 2011), we suspected that the gain in selectivity is primarily driven by hydrophobic interactions (e.g., the amount of hydrophobic protein surface buried upon ligand binding). The general magnitude of the so-called hydrophobic effects obtained from diverse sets of protein-ligand complexes for an additional methyl group is 30 cal/mol × Å2 (Vallone et al., 1998), which is equivalent to a 3.5-fold increase in binding affinity (Bissantz et al., 2010) and roughly on the order of magnitude observed here for the increases and decreases in potency. A detailed examination of the van der Waals interactions in the lowest energy-binding poses of SKA-121 and SKA-31 in our Rosetta models shows that both SKA-121 and SKA-31 make more hydrophobic contacts in the KCa3.1- than the KCa2.3–CaM-BD/CaM interface pocket. While “the back” of both the KCa3.1 and the KCa2.3 interface is well “filled” by both compounds, actually so well that the presence of any water molecules or ions in the interface is highly unlikely after ligand binding, there is an overall greater shape complementarity between the lipophilic binding pocket formed by S372, M368, and M72 and the five-position methyl group of SKA-121, which we postulate accounts for the gain in KCa3.1 selectivity observed with SKA-121.

Taken together, we believe our current study provides an explanation for the generally greater potency of small-molecule KCa channel activators such as SKA-121 on KCa3.1 than on KCa2.3. However, we were surprised by the variance in binding poses observed for SKA-31, its derivatives SKA-121 and SKA-111, EBIO, and NS309. Depending on their substitution pattern and the heteroatoms present in their ring systems, the compounds twisted, turned, and shifted around in the CaM-BD/CaM interface pocket, and often underwent quite individualistic interactions with channel residues. Although these findings are encouraging when considering the possibility of designing and identifying compounds with increased KCa channel subtype selectivity, they are also sobering and caution against transferring structure activity relationship findings between even closely related chemotypes without experimental data. Further structural, modeling, and experimental studies will be needed to reveal the molecular mechanisms of action of small-molecule KCa channel activators in the full-length KCa channel structures.

Authorship Contributions

Participated in research design: Brown, Shim, Zhang, Yarov-Yarovoy, Wulff.

Conducted experiments: Brown, Shim, Yarov-Yarovoy.

Performed data analysis: Brown, Shim, Yarov-Yarovoy.

Wrote or contributed to the writing of the manuscript: Brown, Shim, Zhang, Yarov-Yarovoy, Wulff.

Footnotes

    • Received May 18, 2017.
    • Accepted July 27, 2017.

  • This work was supported by the CounterACT Program, National Institutes of Health Office of the Director [U54NS079202], and the National Institute of Neurologic Disorders and Stroke [R21NS101876].

  • https://doi.org/10.1124/mol.117.109421.

  • ↵Embedded ImageThis article has supplemental material available at molpharm.aspetjournals.org.

Abbreviations

CaM
calmodulin
CaM-BD
calmodulin-binding domain
(−)CM-TMPF
N-{7-[1-(4-chloro-2-methylphenoxy)ethyl]-[1,2,4]triazolo[1,5-a]pyrimidin-2-yl}-N′-methoxy-formamidine)
EBIO
1-ethylbenzimidazolin-2-one
FI
fold increase
KCa
Ca2+-activated K+ channel
KCa2
small conductance Ca2+-activated K+ channel
KCa3.1
intermediate-conductance Ca2+-activated K+ channel
NS309
6,7-dichloro-1H-indole-2,3-dione 3-oxime
NS6180
4-[[3-(trifluoromethyl)phenyl]methyl]-2H-1,4-benzothiazin-3(4H)-one
NS8593
(R)-N-(benzimidazol-2-yl)-1,2,3,4-tetrahydro-1-naphtylamine
pdb
Protein Data Bank
PO
open probability
SKA-111
5-methylnaphtho[1,2-d]thiazol-2-amine
SKA-121
5-methylnaphtho[2,1-d]oxazol-2-amine
SKA-31
naphtho[1,2-d]thiazol-2-ylamine
TRAM-34
1-[(2-chlorophenyl)diphenylmethyl]-1H-pyrazole
UCL-1684
6,10-diaza-3(1,3)8(1,4)-dibenzena-1,5(1,4)-diquinolinacyclodecaphane
UCL-1848
8,14-diaza-1,7(1,4)-diquinolinacyclotetradecaphane
WT
wild-type

References

  1. ↵
    1. Adelman JP,
    2. Maylie J, and
    3. Sah P
    (2012) Small-conductance Ca2+-activated K+ channels: form and function. Annu Rev Physiol 74:245–269.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Barreiro EJ,
    2. Kümmerle AE, and
    3. Fraga CA
    (2011) The methylation effect in medicinal chemistry. Chem Rev 111:5215–5246.
    OpenUrlPubMed
  3. ↵
    1. Barth P,
    2. Schonbrun J, and
    3. Baker D
    (2007) Toward high-resolution prediction and design of transmembrane helical protein structures. Proc Natl Acad Sci USA 104:15682–15687.
    OpenUrlAbstract/FREE Full Text
  4. ↵
    1. Bissantz C,
    2. Kuhn B, and
    3. Stahl M
    (2010) A medicinal chemist’s guide to molecular interactions. J Med Chem 53:5061–5084.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Bonneau R,
    2. Strauss CE,
    3. Rohl CA,
    4. Chivian D,
    5. Bradley P,
    6. Malmström L,
    7. Robertson T, and
    8. Baker D
    (2002) De novo prediction of three-dimensional structures for major protein families. J Mol Biol 322:65–78.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Bruening-Wright A,
    2. Lee WS,
    3. Adelman JP, and
    4. Maylie J
    (2007) Evidence for a deep pore activation gate in small conductance Ca2+-activated K+ channels. J Gen Physiol 130:601–610.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    1. Bruening-Wright A,
    2. Schumacher MA,
    3. Adelman JP, and
    4. Maylie J
    (2002) Localization of the activation gate for small conductance Ca2+-activated K+ channels. J Neurosci 22:6499–6506.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Busse R,
    2. Edwards G,
    3. Félétou M,
    4. Fleming I,
    5. Vanhoutte PM, and
    6. Weston AH
    (2002) EDHF: bringing the concepts together. Trends Pharmacol Sci 23:374–380.
    OpenUrlCrossRefPubMed
  9. ↵
    1. Cao Y,
    2. Dreixler JC,
    3. Roizen JD,
    4. Roberts MT, and
    5. Houamed KM
    (2001) Modulation of recombinant small-conductance Ca(2+)-activated K(+) channels by the muscle relaxant chlorzoxazone and structurally related compounds. J Pharmacol Exp Ther 296:683–689.
    OpenUrlAbstract/FREE Full Text
  10. ↵
    1. Castle NA,
    2. London DO,
    3. Creech C,
    4. Fajloun Z,
    5. Stocker JW, and
    6. Sabatier J-M
    (2003) Maurotoxin: a potent inhibitor of intermediate conductance Ca2+-activated potassium channels. Mol Pharmacol 63:409–418.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Chandy KG,
    2. Fantino E,
    3. Wittekindt O,
    4. Kalman K,
    5. Tong LL,
    6. Ho TH,
    7. Gutman GA,
    8. Crocq MA,
    9. Ganguli R,
    10. Nimgaonkar V, et al.
    (1998) Isolation of a novel potassium channel gene hSKCa3 containing a polymorphic CAG repeat: a candidate for schizophrenia and bipolar disorder? Mol Psychiatry 3:32–37.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Christophersen P and
    2. Wulff H
    (2015) Pharmacological gating modulation of small- and intermediate-conductance Ca(2+)-activated K(+) channels (KCa2.x and KCa3.1). Channels (Austin) 9:336–343.
    OpenUrl
  13. ↵
    1. Coleman N,
    2. Brown BM,
    3. Oliván-Viguera A,
    4. Singh V,
    5. Olmstead MM,
    6. Valero MS,
    7. Köhler R, and
    8. Wulff H
    (2014) New positive Ca2+-activated K+ channel gating modulators with selectivity for KCa3.1. Mol Pharmacol 86:342–357.
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Davis IW and
    2. Baker D
    (2009) RosettaLigand docking with full ligand and receptor flexibility. J Mol Biol 385:381–392.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Davis IW,
    2. Raha K,
    3. Head MS, and
    4. Baker D
    (2009) Blind docking of pharmaceutically relevant compounds using RosettaLigand. Protein Sci 18:1998–2002.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Devor DC,
    2. Singh AK,
    3. Frizzell RA, and
    4. Bridges RJ
    (1996) Modulation of Cl- secretion by benzimidazolones. I. Direct activation of a Ca(2+)-dependent K+ channel. Am J Physiol 271:L775–L784.
    OpenUrl
  17. ↵
    1. Fanger CM,
    2. Ghanshani S,
    3. Logsdon NJ,
    4. Rauer H,
    5. Kalman K,
    6. Zhou J,
    7. Beckingham K,
    8. Chandy KG,
    9. Cahalan MD, and
    10. Aiyar J
    (1999) Calmodulin mediates calcium-dependent activation of the intermediate conductance KCa channel, IKCa1. J Biol Chem 274:5746–5754.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Feske S,
    2. Wulff H, and
    3. Skolnik EY
    (2015) Ion channels in innate and adaptive immunity. Annu Rev Immunol 33:291–353.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Gerlach AC,
    2. Syme CA,
    3. Giltinan L,
    4. Adelman JP, and
    5. Devors DC
    (2001) ATP-dependent activation of the intermediate conductance, Ca2+-activated K+ channel, hIK1, is conferred by a C-terminal domain. J Biol Chem 276:10963–10970.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Hawkins PC and
    2. Nicholls A
    (2012) Conformer generation with OMEGA: learning from the data set and the analysis of failures. J Chem Inf Model 52:2919–2936.
    OpenUrlCrossRefPubMed
  21. ↵
    1. Hawkins PC,
    2. Skillman AG,
    3. Warren GL,
    4. Ellingson BA, and
    5. Stahl MT
    (2010) Conformer generation with OMEGA: algorithm and validation using high quality structures from the Protein Databank and Cambridge Structural Database. J Chem Inf Model 50:572–584.
    OpenUrlCrossRefPubMed
  22. ↵
    1. Hougaard C,
    2. Eriksen BL,
    3. Jørgensen S,
    4. Johansen TH,
    5. Dyhring T,
    6. Madsen LS,
    7. Strøbaek D, and
    8. Christophersen P
    (2007) Selective positive modulation of the SK3 and SK2 subtypes of small conductance Ca2+-activated K+ channels. Br J Pharmacol 151:655–665.
    OpenUrlCrossRefPubMed
  23. ↵
    1. Hougaard C,
    2. Hammami S,
    3. Eriksen BL,
    4. Sørensen US,
    5. Jensen ML,
    6. Strøbæk D, and
    7. Christophersen P
    (2012) Evidence for a common pharmacological interaction site on K(Ca)2 channels providing both selective activation and selective inhibition of the human K(Ca)2.1 subtype. Mol Pharmacol 81:210–219.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Jenkins DP,
    2. Strøbæk D,
    3. Hougaard C,
    4. Jensen ML,
    5. Hummel R,
    6. Sørensen US,
    7. Christophersen P, and
    8. Wulff H
    (2011) Negative gating modulation by (R)-N-(benzimidazol-2-yl)-1,2,3,4-tetrahydro-1-naphthylamine (NS8593) depends on residues in the inner pore vestibule: pharmacological evidence of deep-pore gating of K(Ca)2 channels. Mol Pharmacol 79:899–909.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Jensen BS,
    2. Strobaek D,
    3. Christophersen P,
    4. Jorgensen TD,
    5. Hansen C,
    6. Silahtaroglu A,
    7. Olesen SP, and
    8. Ahring PK
    (1998) Characterization of the cloned human intermediate-conductance Ca2+-activated K+ channel. Am J Physiol 275:C848–C856.
    OpenUrl
  26. ↵
    1. Jones HM,
    2. Bailey MA,
    3. Baty CJ,
    4. Macgregor GG,
    5. Syme CA,
    6. Hamilton KL, and
    7. Devor DC
    (2007) An NH2-terminal multi-basic RKR motif is required for the ATP-dependent regulation of hIK1. Channels (Austin) 1:80–91.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Lam J,
    2. Coleman N,
    3. Garing AL, and
    4. Wulff H
    (2013) The therapeutic potential of small-conductance KCa2 channels in neurodegenerative and psychiatric diseases. Expert Opin Ther Targets 17:1203–1220.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Lappin SC,
    2. Dale TJ,
    3. Brown JT,
    4. Trezise DJ, and
    5. Davies CH
    (2005) Activation of SK channels inhibits epileptiform bursting in hippocampal CA3 neurons. Brain Res 1065:37–46.
    OpenUrlPubMed
  29. ↵
    1. Leung CS,
    2. Leung SS,
    3. Tirado-Rives J, and
    4. Jorgensen WL
    (2012) Methyl effects on protein-ligand binding. J Med Chem 55:4489–4500.
    OpenUrlCrossRefPubMed
  30. ↵
    1. Li W,
    2. Halling DB,
    3. Hall AW, and
    4. Aldrich RW
    (2009) EF hands at the N-lobe of calmodulin are required for both SK channel gating and stable SK-calmodulin interaction. J Gen Physiol 134:281–293.
    OpenUrlAbstract/FREE Full Text
  31. ↵
    1. Logsdon NJ,
    2. Kang J,
    3. Togo JA,
    4. Christian EP, and
    5. Aiyar J
    (1997) A novel gene, hKCa4, encodes the calcium-activated potassium channel in human T lymphocytes. J Biol Chem 272:32723–32726.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Mandell DJ,
    2. Coutsias EA, and
    3. Kortemme T
    (2009) Sub-angstrom accuracy in protein loop reconstruction by robotics-inspired conformational sampling. Nat Methods 6:551–552.
    OpenUrlCrossRefPubMed
  33. ↵
    1. Meiler J and
    2. Baker D
    (2006) ROSETTALIGAND: protein-small molecule docking with full side-chain flexibility. Proteins 65:538–548.
    OpenUrlCrossRefPubMed
  34. ↵
    1. Nguyen HM,
    2. Singh V,
    3. Pressly B,
    4. Jenkins DP,
    5. Wulff H, and
    6. Yarov-Yarovoy V
    (2017) Structural insights into the atomistic mechanisms of action of small molecule inhibitors targeting the KCa3.1 channel pore. Mol Pharmacol 91:392–402.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Pedarzani P,
    2. Mosbacher J,
    3. Rivard A,
    4. Cingolani LA,
    5. Oliver D,
    6. Stocker M,
    7. Adelman JP, and
    8. Fakler B
    (2001) Control of electrical activity in central neurons by modulating the gating of small conductance Ca2+-activated K+ channels. J Biol Chem 276:9762–9769.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Pettersen EF,
    2. Goddard TD,
    3. Huang CC,
    4. Couch GS,
    5. Greenblatt DM,
    6. Meng EC, and
    7. Ferrin TE
    (2004) UCSF Chimera--a visualization system for exploratory research and analysis. J Comput Chem 25:1605–1612.
    OpenUrlCrossRefPubMed
  37. ↵
    1. Rauer H,
    2. Lanigan MD,
    3. Pennington MW,
    4. Aiyar J,
    5. Ghanshani S,
    6. Cahalan MD,
    7. Norton RS, and
    8. Chandy KG
    (2000) Structure-guided transformation of charybdotoxin yields an analog that selectively targets Ca(2+)-activated over voltage-gated K(+) channels. J Biol Chem 275:1201–1208.
    OpenUrlAbstract/FREE Full Text
  38. ↵
    1. Rohl CA,
    2. Strauss CE,
    3. Misura KM, and
    4. Baker D
    (2004) Protein structure prediction using Rosetta. Methods Enzymol 383:66–93.
    OpenUrlCrossRefPubMed
  39. ↵
    1. Sankaranarayanan A,
    2. Raman G,
    3. Busch C,
    4. Schultz T,
    5. Zimin PI,
    6. Hoyer J,
    7. Köhler R, and
    8. Wulff H
    (2009) Naphtho[1,2-d]thiazol-2-ylamine (SKA-31), a new activator of KCa2 and KCa3.1 potassium channels, potentiates the endothelium-derived hyperpolarizing factor response and lowers blood pressure. Mol Pharmacol 75:281–295.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Singh S,
    2. Syme CA,
    3. Singh AK,
    4. Devor DC, and
    5. Bridges RJ
    (2001) Benzimidazolone activators of chloride secretion: potential therapeutics for cystic fibrosis and chronic obstructive pulmonary disease. J Pharmacol Exp Ther 296:600–611.
    OpenUrlAbstract/FREE Full Text
  41. ↵
    1. Stocker JW,
    2. De Franceschi L,
    3. McNaughton-Smith GA,
    4. Corrocher R,
    5. Beuzard Y, and
    6. Brugnara C
    (2003) ICA-17043, a novel Gardos channel blocker, prevents sickled red blood cell dehydration in vitro and in vivo in SAD mice. Blood 101:2412–2418.
    OpenUrlAbstract/FREE Full Text
  42. ↵
    1. Strøbæk D,
    2. Brown DT,
    3. Jenkins DP,
    4. Chen YJ,
    5. Coleman N,
    6. Ando Y,
    7. Chiu P,
    8. Jørgensen S,
    9. Demnitz J,
    10. Wulff H, et al.
    (2013) NS6180, a new K(Ca) 3.1 channel inhibitor prevents T-cell activation and inflammation in a rat model of inflammatory bowel disease. Br J Pharmacol 168:432–444.
    OpenUrlCrossRefPubMed
  43. ↵
    1. Syme CA,
    2. Gerlach AC,
    3. Singh AK, and
    4. Devor DC
    (2000) Pharmacological activation of cloned intermediate- and small-conductance Ca(2+)-activated K(+) channels. Am J Physiol Cell Physiol 278:C570–C581.
    OpenUrlAbstract/FREE Full Text
  44. ↵
    1. Vallone B,
    2. Miele AE,
    3. Vecchini P,
    4. Chiancone E, and
    5. Brunori M
    (1998) Free energy of burying hydrophobic residues in the interface between protein subunits. Proc Natl Acad Sci USA 95:6103–6107.
    OpenUrlAbstract/FREE Full Text
  45. ↵
    1. Wang C,
    2. Bradley P, and
    3. Baker D
    (2007) Protein-protein docking with backbone flexibility. J Mol Biol 373:503–519.
    OpenUrlCrossRefPubMed
  46. ↵
    1. Wei AD,
    2. Gutman GA,
    3. Aldrich R,
    4. Chandy KG,
    5. Grissmer S, and
    6. Wulff H
    (2005) International Union of Pharmacology. LII. Nomenclature and molecular relationships of calcium-activated potassium channels. Pharmacol Rev 57:463–472.
    OpenUrlFREE Full Text
  47. ↵
    1. Wulff H and
    2. Castle NA
    (2010) Therapeutic potential of KCa3.1 blockers: recent advances and promising trends. Expert Rev Clin Pharmacol 3:385–396.
    OpenUrlCrossRefPubMed
  48. ↵
    1. Wulff H,
    2. Gutman GA,
    3. Cahalan MD, and
    4. Chandy KG
    (2001) Delineation of the clotrimazole/TRAM-34 binding site on the intermediate conductance calcium-activated potassium channel, IKCa1. J Biol Chem 276:32040–32045.
    OpenUrlAbstract/FREE Full Text
  49. ↵
    1. Wulff H,
    2. Kolski-Andreaco A,
    3. Sankaranarayanan A,
    4. Sabatier JM, and
    5. Shakkottai V
    (2007) Modulators of small- and intermediate-conductance calcium-activated potassium channels and their therapeutic indications. Curr Med Chem 14:1437–1457.
    OpenUrlCrossRefPubMed
  50. ↵
    1. Wulff H,
    2. Miller MJ,
    3. Hansel W,
    4. Grissmer S,
    5. Cahalan MD, and
    6. Chandy KG
    (2000) Design of a potent and selective inhibitor of the intermediate-conductance Ca2+-activated K+ channel, IKCa1: a potential immunosuppressant. Proc Natl Acad Sci USA 97:8151–8156.
    OpenUrlAbstract/FREE Full Text
  51. ↵
    1. Xia XM,
    2. Fakler B,
    3. Rivard A,
    4. Wayman G,
    5. Johnson-Pais T,
    6. Keen JE,
    7. Ishii T,
    8. Hirschberg B,
    9. Bond CT,
    10. Lutsenko S, et al.
    (1998) Mechanism of calcium gating in small-conductance calcium-activated potassium channels. Nature 395:503–507.
    OpenUrlCrossRefPubMed
  52. ↵
    1. Yarov-Yarovoy V,
    2. Schonbrun J, and
    3. Baker D
    (2006) Multipass membrane protein structure prediction using Rosetta. Proteins 62:1010–1025.
    OpenUrlCrossRefPubMed
  53. ↵
    1. Zhang M,
    2. Pascal JM,
    3. Schumann M,
    4. Armen RS, and
    5. Zhang JF
    (2012) Identification of the functional binding pocket for compounds targeting small-conductance Ca2+-activated potassium channels. Nat Commun 3:1021.
    OpenUrlCrossRefPubMed
  54. ↵
    1. Zhang M,
    2. Pascal JM, and
    3. Zhang JF
    (2013) Unstructured to structured transition of an intrinsically disordered protein peptide in coupling Ca2+-sensing and SK channel activation. Proc Natl Acad Sci USA 110:4828–4833.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Download PDF
Article Alerts
Email Article
Citation Tools

Research ArticleArticle

Molecular Model of the SKA-121 Binding Site

Brandon M. Brown, Heesung Shim, Miao Zhang, Vladimir Yarov-Yarovoy and Heike Wulff
Molecular Pharmacology October 1, 2017, 92 (4) 469-480; DOI: https://doi.org/10.1124/mol.117.109421
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

Advertisement