Toward the Rational Design of Constitutively Active KCa3.1 Mutant Channels

C H A P T E R

T W E N T Y- F O U R

Toward the Rational Design of Constitutively Active KCa3.1 Mutant Channels Line Garneau, He´le`ne Klein, Lucie Parent, and Re´my Sauve´ Contents 1. Introduction 1.1. KCa channels 1.2. KCa3.1 physiological role in health and diseases 1.3. Rationale for the design of constitutively active KCa3.1 channels 2. Production of Constitutively Active KCa3.1 Mutant Channels 2.1. Production of a model structure of the pore region 2.2. Experimental identification of the residues forming the channel pore 2.3. Testing the current leak hypothesis 2.4. Constitutive activity and the energetics of the S6 transmembrane helix 2.5. Are constitutively active KCa3.1 representative of the channel open configuration? 3. Concluding Remarks References

438 438 440 440 441 441 444 446 448 451 454 455

Abstract The Ca2þ activated potassium channel of intermediate conductance KCa3.1 is now emerging as a therapeutic target for a large variety of health disorders. KCa3.1 is a tetrameric membrane protein with each subunit formed of six transmembrane helices (S1–S6). Ca2þ sensitivity is conferred by the Ca2þ binding protein calmodulin (CaM), with the CaM C-lobe constitutively bound to an intracellular domain of the channel C-terminus, located proximal to the membrane and connected to the S6 transmembrane segment. Patch clamp single channel recordings have demonstrated that binding of Ca2þ to CaM allows the channel to transit dose dependently from a nonconducting to an Department of Physiology, Groupe d’e´tude des prote´ines membranaires, Universite´ de Montre´al, Montreal, Canada Methods in Enzymology, Volume 485 ISSN 0076-6879, DOI: 10.1016/S0076-6879(10)85024-4

#

2010 Elsevier Inc. All rights reserved.

437

438

Line Garneau et al.

ion-conducting configuration. Here we present a general strategy to generate KCa3.1 mutant channels that remain in an ion-conducting state in the absence of Ca2þ. Our strategy is first based on the production of a 3D model of the channel pore region, followed by SCAM experiments to confirm that residues along each of the channel S6 transmembrane helix form the channel pore lumen as predicted. In a simple model, constitutive activity can be obtained by removing the steric hindrances inside the channel pore susceptible to prevent ion flow when the channel is in the closed configuration. Using charged MTS reagents and Agþ ions as probes acting on Cys residues engineered in the pore lumen, we found that the S6 transmembrane helices of KCa3.1 cannot form a pore constriction tight enough to prevent ion flow for channels in the closed state. These observations ruled out experimental strategies where constitutive activity would be generated by producing a “leaky” closed channel. A more successful approach consisted however in perturbing the channel open/closed state equilibrium free energy. In particular, we found that substituting the hydrophobic residue V282 in S6 by hydrophilic amino acids could lock the channel in an open-like state, resulting in channels that were ion conducting in the absence of Ca2þ.

1. Introduction 1.1. KCa channels Calcium signaling cascades play a prominent role in a large variety of cellular processes. Calcium activated potassium channels constitute in this regard key effectors during Ca2þ signaling as increases in cystosolic Ca2þ concentration cause an enhanced channel activity resulting in a variation of the membrane potential in both excitable and nonexcitable cells. Based on their single channel conductance, genetic relationship and mechanisms of Ca2þ activation, the eight KCa channels identified so far form two well defined groups. The first group contains the KCa1.1, KCa4.1, KCa4.2, and KCa5.1 channels characterized by rather large unitary conductance (>100 pS). The best studied channel of this group is KCa1.1 (Maxi KCa), which is both voltage and Ca2þ sensitive. Ca2þ sensitivity in this case is linked to specific domains of the channel structure involved in Ca2þ binding. The second group refers to the three small conductance channels, KCa2.1, KCa2.2, and KCa2.3 and to the intermediate conductance KCa3.1 channel. These channels are voltage insensitive, and Ca2þ sensitivity is conferred by the Ca2þ-binding protein calmodulin (CaM) constitutively bound in C-terminus to a membrane-proximal region that is highly conserved among these channels (Khanna et al., 1999). KCa3.1 and KCa2.x present, however, distinct pharmacological profiles with KCa3.1 and KCa2.x, respectively, inhibited by TRAM-34 (Wulff et al., 2000) and apamin (Kohler et al., 1996).

439

Constitutively Active KCa3.1

261 S1 39

S2

S3

S4

S5

203

179

LZ 18

S6

287 NDPK-B

17 312 15

NH3+

– COO

C-CaM 329

Ca2+

MTMR6

Ca2+

P

His

358

PHPT-1 406

LZ

378

372

N-CaM

361

Figure 24.1 Membrane topology of KCa3.1. KCa3.1 is a tetrameric protein with each monomer organized in six transmembrane segments plus a pore region between segments 5 and 6. The channel Ca2þ sensitivity is conferred by calmodulin, with the CaM C-lobe (C-CaM) constitutively bound to the 312–329 segment of the channel Cterminal region. A stretch of 11 amino acids from 361 to 372 is responsible for the Ca2þ-dependent binding of the CaM N-lobe (N-CaM) to the channel. The channel regulation by ATP is mediated in part by NDPK-B and PHPT-1 which phosphorylates and dephosphorylates H358, while the presence of a coil–coil segment at the channel C-terminal end is involved in the binding of the MTMR6 phosphatase. We have shown that AMPK-g1-subunit can interact with a domain extending from 380 to 400 in KCa3.1 C-terminus (Klein et al., 2009). Finally, KCa3.1 contains a 15-RKR-17 motif in N-terminus required for ATP regulation, and two leucine zipper motifs (LZ) in Nand C-termini respectively, critical for channel assembly and trafficking.

Cloning and functional expression of the KCa3.1 channel have revealed that KCa3.1 is a tetrameric membrane protein with each subunit organized in six transmembrane segments S1–S6 with a pore motif between segments 5 (S5) and 6 (S6) (Fig. 24.1). Each subunit contains in C-terminus a Ca2þ-dependent and a Ca2þ-independent CaM binding domain, for an overall stoichiometry of four CaM per channel (Khanna et al., 1999). Recent studies have also shown that internal ATP stimulates the human KCa3.1 channel activity via phosphorylation by the nucleoside diphosphate kinase NDPK-B of a histidine residue at position 358 located within the channel C-terminal region (Srivastava et al., 2006b). The stimulatory action of ATP requires Ca2þ with ATP leading to an apparent increase in Ca2þ sensitivity (Srivastava et al., 2006a). Work from our laboratory has also provided the first evidence that the C-terminal region of KCa3.1 is interacting with the g1-subunit of the metabolic sensing kinase AMPK (Klein et al., 2009), with KCa3.1 activity decreasing in response to AMPK stimulation. Altogether these data support a model where KCa3.1 is part of a multiprotein complex involving several protein kinases.

440

Line Garneau et al.

1.2. KCa3.1 physiological role in health and diseases There is now strong evidence that the KCa3.1 channel plays a prominent role in a large variety of physiological events. For instance, the KCa3.1 blocker TRAM-34 has been documented to suppress acute immune reactions involving memory B and T cells (Wulff et al., 2004). KCa3.1 is also well known to constitute a major determinant to the endotheliumdependent control of vascular tone. Data from several laboratories have established that KCa3.1 activation is an obligatory step to the EDHF (endothelium-derived hyperpolarizing factor) vasodilation process, with TRAM-34 and apamin causing an inhibition of the EDHF-induced vasorelaxation (Feletou and Vanhoutte, 2007). A strong upregulation of KCa3.1 was observed in the mitogenesis of rat fibroblast, vascular smooth muscle cells (Grgic et al., 2005), and cancer cell lines, indicating that KCa3.1 represents an important regulator of cell proliferation (Grgic et al., 2005; Jager et al., 2004; Ouadid-Ahidouch et al., 2004). Notably, the use of the KCa3.1 inhibitor TRAM-34 was found to prevent restenosis, an effect directly related to an upregulated expression of KCa3.1 in coronary artery vascular smooth muscle cells following balloon catheter injury (Kohler et al., 2003). Recent data also demonstrated that KCa3.1 is involved in renal fibroblast proliferation and fibrogenesis (Grgic et al., 2009) suggesting that KCa3.1 may represent a therapeutic target for the treatment of fibrotic kidney disease. Finally, increasing evidence argues for a prominent role of KCa3.1 in Cl secreting epithelial cells by maintaining an electrochemical gradient favorable to Cl and Naþ transepithelial transport. The strong coupling between basolateral Kþ channel activation and apical Cl secretion has led to the proposal of using Kþ channel openers such as DCEBIO (Singh et al., 2001) or 4-chloro-benzo[F]isoquinoline (CBIQ; Szkotak et al., 2004) as therapeutic agents to correct fluid secretion in epithelia presenting ion transport defects as found in cystic fibrosis. In contrast, blockage of KCa3.1 might be beneficial in treating pathological conditions characterized by an excessive fluid secretion. In this regard, the use of the KCa3.1 blocker clotrimazole was documented to normalize salt and water transport in secretory diarrhea (Rufo et al., 1997). Altogether, these data support KCa3.1 as promising therapeutic target for a large variety of health disorders.

1.3. Rationale for the design of constitutively active KCa3.1 channels Investigating the KCa3.1 structures involved in transducing Ca2þ binding into channel opening is essential, not only to understand how the channel actually works, but also to determine how channel activity can be affected by blockers or potentiators. Critical to this process is the identification of

Constitutively Active KCa3.1

441

residues that when mutated lead to channels that are ion conducting in the absence of Ca2þ. These residues may either critically contribute to the energy balance between the channel open/closed configurations, or play a pivotal role in coupling the channel activation gate to the conformational change triggered by the binding of Ca2þ to the CaM/KCa3.1 complex. Identification of these residues thus requires to localize the channel activation gate and/or the structures responsible to maintain the gate in an open state. A study of the residues involved in constitutive activity may also offer the possibility to generate KCa3.1 mutants susceptible to be used in high-throughput screening (HTPS) assays. Testing drugs by fluorescence-based membrane potential measurements represents a valid screening strategy as long as the changes in membrane potential accurately reflect the activity of the channel of interest. In this regard, the study of KCa3.1 is particularly challenging. First, the maximum open probability of KCa3.1 rarely exceeds 0.2 in saturating Ca2þ conditions, so that changes in membrane potential due to KCa3.1 inhibition in resting Ca2þ conditions might not be important enough to be detectable through fluorescence measurements. Second, HTPS cannot exclude indirect effects of the drugs on proteins involved in KCa3.1 regulation. For instance, drugs that will affect intracellular Ca2þ homeostasis or regulate the activity of protein kinases such as AMPK or NDPK-B are expected to modify the KCa3.1 contribution to membrane potential independently of a direct action on the channel itself. To circumvent this problem, one may consider the use of a KCa3.1 channel constitutively active in zero Ca2þ. Besides being insensitive to intracellular Ca2þ fluctuations, this channel would also be independent of ATP-based regulatory mechanisms. An automated selection of channel inhibitors could under these conditions truly reflect an effect of the drug on the channel itself, and not on some secondary mechanisms. To be truly applicable, however, this approach requires that the structure of the channel constitutively active be representative of the wild-type channel in the open configuration. Here we present the global strategy that was implemented to produce KCa3.1 mutant channels that showed constitutive activation properties.

2. Production of Constitutively Active KCa3.1 Mutant Channels 2.1. Production of a model structure of the pore region The rational design of a constitutively active KCa3.1 mutant first requires to identify some of the channel key structural features. Unfortunately, the 3D structure of KCa3.1 is currently unknown. A 3D representation of KCa3.1 pore region could, however, be generated by applying a comparative

442

Line Garneau et al.

modeling approach where segments of the channel amino acid sequence were translated into 3D structural data (Baker and Sali, 2001; Sali and Blundell, 1993). Comparative modeling, or homology modeling, is usually based on a number of steps: (1) identifying a known structure (the template structure) with an amino acid sequence homologous to the sequence of the protein to be modeled (the target sequence); (2) aligning the template sequence and secondary structure to the target amino acid sequence; (3) building model structures using a specialized software such as MODELLER; and (4) testing the model. Identification of suitable template structures can be obtained using the computational tools provided by specialized servers such as SAM-T08 (http://compbio.soe.ucsc.edu/SAM_T08/T08query.html) and/or T-TASSER (http://zhanglab.ccmb.med.umich.edu/). This procedure is usually complemented by a comparative sequence analysis with T-COFFEE and MUSCLE as to fine tune the sequence alignments that will serve as input files for MODELLER. Using such a procedure for KCa3.1, we identified the Kv1.2 (Long et al., 2005; PDB:2a79) and MlotiK1 (Clayton et al., 2008; PDB:3BEH) potassium channel crystal structures as suitable templates to generate a 3D representation of the open/closed KCa3.1 pore region (Fig. 24.2). MlotiK1 is a voltage insensitive internal ligand-gated channel with six transmembrane helices per monomer. The MlotiK1structure is representative of a ligand-gated channel in the closed configuration, whereas the voltage-gated Kv1.2 channel structure is more representative of a channel in the open state. Performing automated homology modeling with MODELLER V9.3 (Sali and Blundell, 1993) requires the initial production of a large number of models for each template. We routinely set the number of models to 150, and the best models are selected based on the value of the objective function (roughly related to the energy of the model) provided by MODELLER, together with the RMS deviation computed for the Ca of the model structure backbone relative to the template. The overall structural quality of the top five model 3D structures can be checked by PROCHECK (Laskowski et al., 1993) so that aberrant structural features (rotation angles, bond lengths, etc.) can be user adjusted during the procedure. When applied to KCa3.1, the models derived from Kv1.2 and MlotiK1 predicted that the residues V275, T278, A279, V282, and A286 of the S6 transmembrane segment should be lining the channel pore with residues C276, C277, L280, and L281 oriented opposite to the pore lumen (Fig. 24.2). An analysis of the model generated from the MlotiK1 structure indicated furthermore that the narrowest segment of the conduction pathway should be located at the level of the C-terminal end of the four S6 transmembrane helices (A282–A286). This region is predicted to line with the bundle crossing domain of the inverted tepee structure originally described for the KcsA channel (Doyle et al., 1998). More importantly, the bundle crossing region has been proposed to play a prominent role in controlling ion flow in several Kþ channels. For instance,

443

Constitutively Active KCa3.1

A

KCa3.1 closed (MlotiK1)

B

KCa3.1 closed (MlotiK1) Pore helix

V275 S6 S5

V282

S4–S5 linker

C

KCa3.1 open (Kv1.2)

D

KCa3.1 open (Kv1.2)

Pore helix

S6 S5

V275 V282 S4–S5 linker

Figure 24.2 Model structures of KCa3.1. A and C. Global side view representations of the closed (A) and open (C) KCa3.1 embedded in a lipid bilayer (green) separating two electrolyte solutions containing 150 mM KCl. The closed KCa3.1 3D representation was obtained by homology modeling using the MlotiK1 structure (PDB:3BEH) as template, whereas the open representation is based on the Kv1.2 channel structure (PDB:2a79). Both models include the S4–S5 linker plus the S5–S6 pore region of the channel. (B and D) Detailed view of the pore region with only two monomers represented for clarity. Turns colored in blue refer to residues predicted to be facing the pore (V275, T278, A279, V282, and A286). These predictions were confirmed in three SCAM analyses produced by our laboratory. Templates selected with I-TASSER and SAM-T08 servers. CPK representation of Kþ ions in purple. Representation by DS Visualizer.

a major contribution of the bundle crossing region to channel gating was hypothesized from the KirBac1.1 channel structure, where semiconserved phenylalanine residues at the C-terminal end of the inner helices (S6 in KCa3.1) appeared to clash, suggesting that they form a hydrophobic gate capable to block permeation of Kþ ions (Kuo et al., 2003). Similarly,

444

Line Garneau et al.

Cd2þ-dependent block experiments on the voltage activated Shaker channel have revealed that the residue V474 equivalent to V282 in KCa3.1, likely contributes to the formation of a very tight constriction site in the Shaker pore (Webster et al., 2004). It follows that constitutive activation could in principle be obtained by reducing steric hindrance at the level of the bundle crossing region, so that ions would still diffuse through the channel pore despite the channel being in a closed configuration. This is the “leaky” closed state hypothesis for constitutive activity.

2.2. Experimental identification of the residues forming the channel pore The first step to determine whether relieving steric hindrance at the bundle crossing region in KCa3.1 can lead to the formation of an ion-conducting channel in zero Ca2þ is to establish experimentally the nature of the residues predicted to be lining the channel pore. Predictions derived from homology modeling can formally be tested using a method termed substituted-cysteine accessibility method (SCAM) which allows site-selective modification of accessible cysteine in a protein. Each residue of the S6 transmembrane segment was mutated one at a time to a cysteine (Cys) and the effects on channel activity of small, charged, sulfhydryl-specific reagents measured in patch clamp experiments. Cysteine residues engineered along the S6 segment can either be located in a water-accessible environment or buried into the protein. Since the reaction of sulfhydryl-specific reagents such as MTS (methanethiosulfonate) is 109 slower for the protonated than deprotonated form of Cys (Karlin and Akabas, 1998), one expects that only Cys residues exposed to a water filled channel pore will be modified by MTS compared to Cys exposed to a nonaqueous environment. In addition, the irreversible binding of a charged sulfhydryl reagent to a Cys facing the channel pore is likely to inhibit ion flow due to electrostatic interactions and steric hindrance as well. MTSETþ ([2-(trimethylammonium)ethyl] methanethiosulfonate) is a specific sulfhydryl reagent predicted to fit into a cylinder of 2.9 radius and 9 A˚ in length when undergoing an alltrans configuration (without the hydrated shell). Because it is positively charged at neutral pH and poorly liposoluble, this molecule is generally considered to be an excellent probe to investigate the structural features of cationic channel pore. The smaller positively charged MTS reagent ˚ radius  9 A ˚ long) MTSEAþ (aminoethyl methanethiosulfonate: 2.4 A has also been extensively used in SCAM type experiments, but with a pKa > 8.5, 6% of the molecules remains on the average in a neutral form. A neutral molecule could in principal diffuse through the membrane and modify Cys residues located transmembrane to the side of application. This problem can be circumscribed by adding exogenous Cys (1 mM ) to the solution bathing the membrane surface opposite to the side of MTSEAþ

Constitutively Active KCa3.1

445

application (Holmgren et al., 1996). It is recommended finally to use MTS reagents at concentrations lower than 1 mM, to avoid nonspecific channel inhibition coming from overcrowding the channel pore. However, low MTS concentrations (10 mM ) often result in modification time constants of the order of minutes, thus requiring highly stable current recordings with no detectable rundown. The experimental procedure underlying this type of SCAM experiments consists essentially in patch clamp recordings of KCa3.1 channels in the inside-out configuration where the MTS reagent is applied internally by means of a fast solution change system (RSC-160, BioLogic, Grenoble, France). We found that, for most applications, an exchange time less than 30 ms was sufficient to get a reliable measurement of the current change initiated following MTS application. By fitting the time course of the current variation to a single exponential function, the modification rate of the target Cys by a given reagent can be estimated using: modification rate (M 1 s 1) ¼ 1/(MTS concentration (in M )  time constant of the current change (in s)). The bath and patch pipette solutions usually contained (in mM ) 200 K2SO4, 1.8 MgCl2, 0.025 CaCl2, 25 HEPES, buffered at pH 7.4 with KOH. A high Kþ concentration maximize the signal to noise ratio in unitary current recordings and the use of sulfate salts enables to chelate contaminant divalent cations such as Ba2þ (maximum free Ba2þ concentration: 0.5 nM in 200 mM K2SO4) which may otherwise block KCa3.1 by interacting directly with the channel selectivity filter. A sulfate salt also minimizes the contribution coming from endogenous Ca2þ-activated Cl channels in experiments where KCa3.1 is expressed in Xenopus oocytes. Site-directed mutagenesis of KCa3.1 is routinely performed using the QuickChange Site-Directed Mutagenesis kit (Stratagene). Oocytes (stage V or VI) used for channel expression are obtained from Xenopus laevis frogs anaesthetized with 3-aminobenzoic acid ethyl ester. The follicular layer is removed by incubating the oocytes in a Ca2þ-free Barth’s solution containing collagenase (1.6 mg/ml; Sigma-Aldrich) for 60 min. Defolliculated oocytes are stored at 18  C in Barth’s solution supplemented with 5% horse serum, 2.5 mM Na-pyruvate, 100 U/ml penicillin, 0.1 mg/ml kanamycin, and 0.1 mg/ml streptomycin. The Barth’s solution contains (in mM ) 88 NaCl, 3 KCl, 0.82 MgSO4, 0.41 CaCl2, 0.33 Ca(NO3)2, and 5 HEPES (pH 7.6). Oocytes are patched 3–5 days after injection of 0.1-1 ng of the cDNA coding for KCa3.1. Prior to patch clamping, defolliculated oocytes are bathed in a hyperosmotic solution containing (in mM) 250 KCl, 1 MgSO4, 1 EGTA, 50 sucrose, and 10 HEPES buffered at pH 7.4 with KOH. The vitelline membrane is then peeled off using fine forceps, and the oocyte is transferred to a superfusion chamber for patch clamp measurements. This procedure was used to map the pore structure of the open KCa3.1 channel. We found that MTSETþ applied internally caused a total

446

Line Garneau et al.

inhibition of the V275C, T278C, and V282C mutants and partially blocked the A279C and V284C channels. In contrast, MTSETþ initiated a strong channel activation of the A283C and A286C mutant channels. These results are in agreement with the proposed model obtained by homology where V275, T278, A279, and V282 are lining the channel pore in the open state while supporting a bundle crossing region of KCa3.1 extending from V282 to A286 (Klein et al., 2007).

2.3. Testing the current leak hypothesis Central to the current leak hypothesis is the proposal that in the closed state, the channel bundle crossing region should form a seal tight enough to prevent ion flow through the channel pore. Removing steric constraints at this site should consequently lead to the formation of a channel that is ion conducting in the absence of Ca2þ (constitutive activity). The contribution of the S6 segment in the bundle crossing region to the regulation of ion flow can formally be tested by measuring to what extent sulfhydryl-specific reagents of various sizes can access a Cys residue engineered deep in the channel pore when the channel is in the closed configuration. The choice of the probe is crucial in this type of experiments. Because of its size, MTSETþ is not truly representative of a Kþ ion in solution. In fact, the bundle crossing region could be impermeable to MTSETþ for the closed channel, but still be permeable to Kþ ions under identical conditions. MTSEAþ is of ˚ diameter compared to 5.8 A ˚ ) but this smaller size than MTSETþ (4.8 A reagent may lead to false positive results as it can exist in a neutral protonated form and thus access a target Cys deep in the channel pore by diffusing directly through the membrane. Better results can be obtained using Agþ as thiol modifying agent. As mentioned by Lu and Miller (1995), Agþ constitutes an excellent probe to study Kþ channels as both ions are very ˚ for similar in size with a van der Waals radius of 1.52 A˚ for Kþ and 1.29 A þ þ Ag (Marcus, 1988; Shannon, 1976). Ag is, however, highly reactive, so that Agþ solutions need to be prepared by adding AgNO3 to strictly Cl free solutions. In addition, the free Agþ concentration needs to be stabilized using a chelating agent such as EDTA at high concentrations. We found that 60 mM EDTA gave reproducible results. As the diffusion limited modification rate of Cys by Agþ is of the order of 108 M 1 s 1, Agþ should be used at nanomolar concentrations to yield time constant of Cys modification within the seconds range. The free Agþ concentration can be calculated with programs such as Eqcal (Biosoft, Cambridge, UK). Finally, the target Cys should be engineered deep in the channel pore, well above the bundle crossing region, close to the selectivity filter. Our model of the KCa3.1 region predicts that the Val at 275 should be located in proximity of the selectivity filter, facing the channel central cavity. In accordance with this model, we found that the V275C mutant in the open state can be

447

Constitutively Active KCa3.1

specifically blocked through the irreversible binding of sulfhydryl reagents such as MTSEAþ or MTSETþ (Klein et al., 2007). This experimental procedure was applied to our study of the KCa3.1 bundle crossing region. In these experiments, EDTA, EDTA þ Agþ, EDTA solutions were applied repetitively in 1.5 s pulses each separated by a test perfusion with a solution containing 25 mM Ca2þ. Modification rates were calculated from the current inhibition curve obtained in response to cumulative Agþ applications. Our results showed that Cys residues engineered inside the channel central cavity at position 275 were readily accessible to Agþ applied internally (7 nM) with the channel in the closed configuration (see Fig. 24.3). In contrast, larger molecules such as MTSETþ ˚ diameter) showed a 104-fold difference in accessibility between the (5.8 A channel closed and open configurations (Fig. 24.6; Klein et al., 2007). These observations demonstrate that the bundle crossing region for the closed KCa3.1 channel is permeable to Kþ ions, but impermeable to larger molecules such as MTSETþ. This proposal is in line with Cys mutagenesis data from cyclic nucleotide-gated channels suggesting that the inner helices may form a constriction at the C-terminal end of the channel pore tight enough to restrict the accessibility of reagents larger than Kþ to the channel cavity, but nonobstructive to Kþ ion flow (Flynn and Zagotta, 2001, 2003; Xiao et al., 2003). Similar conclusions supported by SCAM data were also

KCa3.1 V275C

7 nM Ag + 60 mM EDTA

I0 0

ICa

40 pA 10 s

I (pA)

–50 –100 –150 –200 –250

0

1

2 3 4 Time (s)

5

6

Figure 24.3 Evidence against the leaky channel model for constitutive activation. Inside-out patch clamp recording illustrating the action of Agþ (7 nM) on the closed V275C channel. Agþ was applied for 0.5 s during a 3-s perfusion period with a Ca2þ free solution containing (in mM ): 150 KMES, 60 EDTA, 10 HEPES, pH 7.3, with 25 mM AgNO3 for a free Agþ concentration of 7 nM. The inhibitory effect of a 0.5s Agþ application was estimated from the current intensity recorded after replacing the zero Ca2þ solution by a solution containing 25 mM Ca2þ (test current). The time dependent variation of the test currents obtained from the repetitive application of Agþ at 0.2 Hz is illustrated in the inserted panel on the right. These results argue against the presence along the channel pore of steric constraints that would impair Kþ ion flow when the channel is in the closed configuration. Originally published in Garneau et al. (2009). # The American Society for Biochemistry and Molecular Biology.

448

Line Garneau et al.

derived from studies performed on the KCa2.2 and Kir2.1 channels (Bruening-Wright et al., 2002; Xiao et al., 2003). It follows that constitutively activated KCa3.1 channels are not likely to be generated by simply removing steric constraints at the bundle crossing site, since this site does not constitute an active barrier controlling Kþ ion flow in this case.

2.4. Constitutive activity and the energetics of the S6 transmembrane helix The exact molecular mechanism underlying KCa3.1 opening in response to Ca2þ binding to the CaM/KCa3.1 complex remains to be elucidated. This point is crucial to the design of a constitutively active KCa3.1 channel. Important structural information pertinent to channel gating was however obtained through the crystallization of CaM bound to the rat KCa2.2-CaM binding domain in the presence of Ca2þ (Schumacher et al., 2001). These results suggested a large-scale conformational rearrangement taking place in the presence of Ca2þ where the N-lobe of CaM binds to a segment in the C-terminus of an adjacent monomer, resulting in a dimerization of contiguous subunits within the channel structure. This rearrangement would in turn lead to a rotation/translation of the associated S6 transmembrane helix and to the opening of the ion-conducting pore (Maylie et al., 2004; Schumacher et al., 2001, 2004; Wissmann et al., 2002). Added to our Agþ ion-based results suggesting an active channel gate at the level of selectivity filter region, this model argues for a crucial role of S6 movements to the KCa3.1 opening process. In fact, the difference in free energy between the channel closed and open state is likely to be governed by the changes in free energy coming from S6 residues making different contacts with their surrounding milieu when the channel is in the open compared to the closed state configuration. To determine to what extent the interactions between the side chain of a residue along S6 with its atomic environment contributes to the difference in free energy between the channel open and closed state, one can minimize these interactions by replacing the residue of interest by a Gly (Garneau et al., 2009). As constitutively active channel mutants will be ion conducting in zero Ca2þ, it is also essential to establish a reference current level corresponding to a non ion-conducting channel. This problem can be circumvented by performing Gly substitutions on the V275C channel, so that MTSEAþ or MTSETþ could be used as specific irreversible blockers of the channel pore. Under these conditions, constitutive activity can be estimated by measuring the ratio R ¼ (IEGTA  I0)/(ICa  I0), where IEGTA and ICa refer, respectively, to the current measured in zero and 25 mM internal free Ca2þ, and I0 the current level following channel block by MTSEAþ. For non constitutively active channels, we expect IEGTA ¼ I0 for R ¼ 0 as the current level in zero Ca2þ will correspond to the

Constitutively Active KCa3.1

449

background current obtained following a total block of the channel pore. Values of R > 0 would be indicative of a channel where IEGTA > I0, thus indicating that the current level obtained in the absence of agonist does not correspond to the current level expected for a nonconducting channel. Using this paradigm, we found that the V275C-V282G and V275CA279G mutants led to R values equal to 1.0  0.02 (n ¼ 3) and 0.2  0.02 (n ¼ 3), respectively, indicating that V275C-V282G or V275C-A279G were ion conducting in zero Ca2þ conditions (Fig. 24.4). Other double mutant channels generated for residues in S6 extending from C276 to A286 did not show constitutive activation properties with R values of 0 (Fig. 24.4D). A change in the channel open/closed equilibrium energy leading to constitutive activity may result either from a stabilization of the channel open configuration, and/or a destabilization of the channel closed configuration. A straightforward approach useful to discriminate between these two mechanisms consists in measuring the channel mean open and closed time at the single channel level. In conditions where N independent channels are present, the channel mean open time htoi can formally be obtained from htoi ¼ hto(r)i[Po(N  r) þ r(1  Po)] / (1  Po) where hto(r)i is the mean open time when r channels are simultaneously open and Po the channel open probability. Po can be calculated from current amplitude histograms assuming that the probability of having ‘r’ channels open simultaneously among N, Po(r), obeys a binomial distribution (Po(r) ¼ N!/[(N  r)!r!] Por(1  Po)N  r ). We found through a single channel analysis of the V275C-A279G mutant that the channel open probability increases from 0.5 at [Ca2þ]i < 0.1 nM to 0.97 in 25 mM Ca2þ, an effect essentially attributable to a 15- to 30-fold increase and 2- to 3-fold decrease of the channel mean open and closed times, respectively. Such a behavior contrasts with the effect of Ca2þ on the wild-type KCa3.1 channel where increasing the internal Ca2þ concentration causes a strong reduction of the channel mean closed time with the channel mean open time remaining rather unchanged. This analysis thus supports a model whereby constitutive activity is obtained by essentially locking the channel in an open configuration (Garneau et al., 2009). Without a detailed description of the microenvironment surrounding the V282 and/or A279 residues in the closed and open state, their respective contribution to the free energy barriers governing the closed to open state transitions is doomed to remain largely undetermined. Hydrophobic interactions have been, however, identified as key determinants to the open/closed equilibrium energy in several ion channels. For instance, constitutively active channels were generated by replacing through yeast screening analysis, the residue V188 (V282 in KCa3.1) in GIRK2 (Yi et al., 2001) by more hydrophilic amino acids. As discussed by Karplus (1997), a quantification in terms of generic hydrophobic effects of the energy difference governing dynamic changes in protein conformations is probably valid for hydrophobic residues only.

450

Line Garneau et al.

A

KCa3.1 V275C

B

I0

KCa3.1 V275C-A279G EGTA MTSEA+

I0 ICa 100 pA

MTSEA+ EGTA

IEGTA

100 s

100 pA 2 s 50 pA

C

KCa3.1 V275C-V282G

I0

25 s ICa

D

KCa3.1 V275C-A286G

EGTA

EGTA

MTSEA+

MTSEA+ I0

ICa IEGTA 25 s

40 pA 10 s 500 pA

ICa

Figure 24.4 Glycine scan along the S6 segment below the Gly hinge at 274. Inside-out recordings of inward currents illustrating the effect of MTSEAþ on the channel mutants. Recordings obtained in symmetrical 200 mM K2SO4 conditions at Vm ¼  60 mV. The symbol ICa refers to the current level in 25 mM Ca2þ, IEGTA to the current level in 1 mM EGTA, and I0 to the current level following inhibition by MTSEAþ. Under these experimental conditions, Ca2þ-activated Kþ current are represented as inwardly directed currents relative to the zero current level I0. (A) Control experiment illustrating the effect of internal MTSEAþ (1 mM) application on the closed V275C channel. MTSEAþ was applied for 3 s during 5.5 s pulses in zero Ca2þ at a frequency of 0.1 Hz. The accessibility to MTSEAþ of cysteines in the channel cavity at position 275 for the closed V275C channel was estimated from the time constant of inhibition of the test inward currents measured in 25 mM internal Ca2þ at the end of each pulse. The successive applications of MTSEAþ in zero Ca2þ did not result in this case in a gradual decrease of the I0 current, but the clear inhibition of the test currents in 25 mM Ca2þ confirms binding of MTSEAþ to the cysteines at 275. (B) Effect of MTSEAþ applied internally on the V275C–A279G double mutant. This current record shows that internal application of MTSEAþ (1 mM) in zero Ca2þ resulted in a strong inward current inhibition for a modification rate of 297  5 M 1 s 1 (n ¼ 3), in support of an ion-conducting conformation in zero Ca2þ. (C) Response of V275C– V282G to MTSEAþ. As seen, the addition of MTSEAþ (1 mM) in zero Ca2þ caused a strong inward current inhibition for a modification rate of 100  15 M 1s 1 (n ¼ 13) confirming that the V275C–V282G mutant was conducting in the absence of Ca2þ. (D) An identical perfusion protocol applied to the V275C–A286G mutant failed to provide evidence of an MTSEAþ-induced inhibition of I0, in accordance with V275C–A286G being nonconductive in the absence of Ca2þ. The observation that channel activity could not be recovered by the addition of Ca2þ following MTSEAþ exposure confirmed that MTSEAþ had access to the cysteines in the channel cavity in zero Ca2þ. Originally published in Garneau et al. (2009). # The American Society for Biochemistry and Molecular Biology.

Constitutively Active KCa3.1

451

The energetics associated to the transfer of a polar or charged residue from a water to a hydrophobic environment require the introduction of additional factors such as H-bound, dipole–dipole and/or Coulombic interactions, which critically depend on the residue microenvironment. Such structural factors remain still ill defined for KCa3.1. If, however, hydrophobic effects constitute the dominant force driving V282 and/or A279 from an aqueous to a nonaqueous environment upon channel closure, there should be a strong correlation between constitutive activity and the free energy of solvation at these particular sites. Selecting residues differing in hydrophobicity relative to Val and/or Ala largely depends on the hydrophobicity scale chosen. In the pure hydrophobicity scale proposed by Karplus, polar atoms of residues are ignored, so that hydrophobic and environment dependent contributions to hydrophobic effects are separated (Karplus, 1997). This scale predicts that the substitution of V282 (3.38 kcal/mol) by residues such as Ala (2.15 kcal/mol), Gln (1.63 kcal/mol), Ser (1.4 kcal/mol), and Gly (1.18 kcal/mol) should lead to a reduced contribution of hydrophobic interactions to the KCa3.1 equilibrium energy, while the substitutions Ile (3.38 kcal/mol) or Leu (4.10 kcal/mol) would be equivalent to Val in terms of hydrophobic effects. By systematically varying the hydrophobicity of the residue at position 282, we found a strong correlation between free energy of solvation and constitutive activity, with residues characterized by a hydrophobic energy for side chain burial less than 1.2 kcal/mol compared to Val more likely to lead to constitutively active channels (Garneau et al., 2009; see Fig. 24.5). This observation truly supports a strategy whereby constitutively active KCa3.1 channels can be generated by decreasing the hydrophobicity of the residues at position 282 in S6, thus stabilizing the channel open configuration. The same approach applied to the A279 residue failed, however, to produce A279 mutants constitutively active, despite A279G being active in zero Ca2þ. This observation points toward either the presence of a side chain at 279 causing systematically a destabilization of the channel open state, or to an effect directly linked to the special structural properties of the Gly residue, likely to be due to an increase in flexibility of the S6 segment at this site.

2.5. Are constitutively active KCa3.1 representative of the channel open configuration? The question remains to what extent the constitutively active V275C-A279G and V275C-V282G mutant channels are representative of the wild-type KCa3.1 in the open state. Formally, the pore structure of the constitutively active V275C-A279G and V275C-V282G channels can be probed by measuring the accessibility of MTS reagents to cysteine residues engineered at position 275. We showed in a previous work that the modification rate of the ˚ diameter) is poorly state dependent, in V275C mutant by MTSEAþ (4.8 A þ contrast to MTSET which yielded modification rates at least 103-fold slower

452

Line Garneau et al.

A

KCa 3.1 V275C-V282L

B

KCa 3.1 V275C-V282D 1mM EGTA

I0

1mM MTSEA+ I0 ICa MTSEA+ EGTA 100 pA 1s

C

KCa 3.1 V275C-V282S 1mM EGTA 1mM MTSEA+

I0

10 pA 25 s IEGTA

D

KCa 3.1 V275C-V282X Constitutively active 10 8 6 Asp 4

Asp(p) IEGTA ICa

Cys 25 s

Gly

IEGTA/ICa

100 pA 25 s

1

Ser

100 pA Ala Val 0 –2.5 –2.0 –1.5 –1.0 –0.5 0.0 ΔΔG (kcal/mole) Gln

IIe 0.5

Leu 1.0

Figure 24.5 Inside-out current recordings illustrating the effect on constitutive activation of mutating V282 by residues of different size and/or hydrophobicity. Inward current recordings performed in symmetrical 200 mM K2SO4 conditions at Vm ¼  60 mV. The symbols ICa, IEGTA and I0, refer to the current recorded at saturating 25 mM Ca2þ, zero Ca2þ or following inhibition with 1 mM MTSEAþ. Ca2þ-activated Kþ currents are represented as inward currents relative to the zero current level I0. The substitution V282L (A) did not result in a constitutively active channel as demonstrated by the absence of current variations in zero Ca2þ (I0 ¼ IEGTA) despite a progressive block of the test inward currents in 25 mM Ca2þ by MTSEAþ. Application of MTSEAþ in zero Ca2þ (EGTA) caused, however, a clear current inhibition with the V275C–V282D (B) and V275C–V282S (C) channel mutants indicating that the V282D and V282S mutations successfully led to constitutive activation. Notably, in contrast to V275C–V282S, the inward current measured with the V275C–V282D mutant was higher in zero Ca2þ (IEGTA) than in 25 mM Ca2þ conditions (ICa). These results are summarized in (D) illustrating the correlation between constitutive activation and hydrophobic energy for side chain burial. Energies were taken from Karplus (1997) and expressed relative to Val. This analysis suggests that residues with a hydrophobic energy for side chain burial less than 1.2 kcal/mol compared to Val are more likely to lead to constitutively active channels. Asp (p) refers to the predicted IEGTA/ICa for V275C–V282D when the channel open probability in 25 mM is corrected for the blocking action of Ca2þ due to Ca2þ binding to the channel selectivity filter. Originally published in Garneau et al. (2009). # The American Society for Biochemistry and Molecular Biology.

in zero than in saturating Ca2þ conditions (Klein et al., 2007). The accessibility of cysteine residues at 275 to MTSETþ can thus be used to assess the conformational state of the open and closed KCa3.1 pore structures. If the S6

453

Constitutively Active KCa3.1

segment of the V275C-A279G and V275C-V282G mutants moves minimally in response of Ca2þ binding to the CaM/KCa3.1 complex, we expect little difference in MTSETþ accessibility for cysteines at 275 with and without Ca2þ. Conversely, if the binding of Ca2þ to the CaM/KCa3.1 complex still induces a movement of the S6 segment, we expect the modification rate by MTSETþ of the cysteine residue at 275 to be Ca2þ dependent despite little effect on channel open probability. Finally, if the rates of modification by MTSETþ of the V275C-A279G and V275C-V282G channels in zero Ca2þ correspond to the rates measured for the open V275C mutant, we will conclude that the pore structure of the constitutive active state of these double mutants is structurally equivalent to the KCa3.1 open configuration. The results of these experiments are summarized in Fig. 24.6. Modification rates were calculated as previously described. Significant differences (p < 0.05) were seen between the modification rates by MTSETþ measured with and without Ca2þ for the V275C channel and for the two constitutively active mutant channels. This observation indicates that despite

Ca2+ EGTA

100

10

1

n=4

n=5

n=4

n=2

n=4

0.1 n=3

MTSET+ modification rate (M–1s–1)

1000

0.01 V275C

V275C–A279G

V275C–V282G

Figure 24.6 Bar graph illustrating the state dependent accessibility to MTSETþ of cysteine residues at position 275. p values of less than 0.05 and 0.0005 are represented as * and ***, respectively. The modification rates measured for the V275C channel and the two V275C–V282G and V275C–A279G mutants differed significantly (p < 0.05) with and without Ca2þ. However, whereas the modification rates measured in zero and 25 mM Ca2þ for the V275C channel differed by 103- to 104-fold, this difference is reduced to less than 10-fold for the constitutively active channels. This effect is attributable to the modification rates in zero Ca2þ, which appeared 50–100 times faster for the constitutively active mutants compared to the V275C control channel. These observations suggest that the pore structure of the V275C–V282G and V275C–A279G mutants in zero Ca2þ better approximates the V275C open than closed configurations. Originally published in Garneau et al. (2009). # The American Society for Biochemistry and Molecular Biology.

454

Line Garneau et al.

constitutive activity both mutant channels remained Ca2þ sensitive. However, whereas the modification rates by MTSETþ measured in zero and 25 mM Ca2þ for the V275C channel differed by 103- to 104-fold, this difference was reduced to less than 10-fold for the constitutively active channels, an effect attributable to the modification rates in zero Ca2þ which appeared 50–100 times faster for the constitutively active mutants compared to the V275C control channel. These results strongly suggest that mutating the residues at positions 279 and 282 into Gly affects the channel geometry in zero Ca2þ so that molecules such as MTSETþ with a van der Waals diameter ˚ have now a greater access to the channel cavity comparatively to the of 5.8 A closed V275C mutant. Finally, our results showed that the accessibility of MTSETþ to V275C in 25 mM Ca2þ is significantly slower (p < 0.05) for the V275C-V282G mutant relative to V275C, an indication that the geometry of the constitutively active state is not totally equivalent to the wild-type open state. Altogether these observations argue for a structure of the constitutively active KCa3.1 that is relatively close to the channel open state without, however, being totally equivalent.

3. Concluding Remarks We have presented a general strategy by which constitutively active KCa3.1 channels can be generated. Our strategy includes (1) the production of a 3D model of the channel pore region coupled to SCAM experiments to identify residues in the S6 transmembrane segment facing the channel pore, (2) a Gly scan of the S6 segment to determine to what extent the interactions between the side chains of key residues along S6 with their surrounding milieu contribute to the open/closed KCa3.1 equilibrium energy, and (3) a perturbation of the channel open/closed equilibrium by varying the contribution of hydrophobic effects to the energy balance leading to constitutive activity. The procedure described in step 1 is essential to establish if constitutive activity cannot be generated by simply removing steric constraints along the channel pore so that the channel would be ion conducting in the closed state. This mechanism does not seem applicable to KCa3.1. In contrast, perturbing the hydrophobic energy of the S6 segment by substituting the Val at position 282 by more hydrophilic amino acids was sufficient to lock the channel in an open-like state. The production of constitutively active mutants thus appears to be tightly linked to the energy balance between the channel open and closed states. In this regard, the results obtained with KCa3.1 are representative of a general mechanism already documented for the Shaker, GIRK2, and KIR3.1/KIR3.4 channels, where constitutive activity is strongly correlated to the energetics of the channel S6 transmembrane segment.

Constitutively Active KCa3.1

455

REFERENCES Baker, D., and Sali, A. (2001). Protein structure prediction and structural genomics. Science 294, 93–96. Bruening-Wright, A., Schumacher, M. A., Adelman, J. P., and Maylie, J. (2002). Localization of the activation gate for small conductance Ca2þ-activated Kþ channels. J. Neurosci. 22, 6499–6506. Clayton, G. M., Altieri, S., Heginbotham, L., Unger, V. M., and Morais-Cabral, J. H. (2008). Structure of the transmembrane regions of a bacterial cyclic nucleotide-regulated channel. Proc. Natl. Acad. Sci. USA 105, 1511–1515. Doyle, D. A., Cabral, J. M., Pfuetzner, R. A., Kuo, A., Gulbis, J. M., Cohen, S. L., Chait, B. T., and MacKinnon, R. (1998). The structure of the potassium channel: Molecular basis of Kþ conduction and selectivity. Science 280, 69–77. Feletou, M., and Vanhoutte, P. M. (2007). Endothelium-dependent hyperpolarizations: Past beliefs and present facts. Ann. Med. 39, 495–516. Flynn, G. E., and Zagotta, W. N. (2001). Conformational changes in S6 coupled to the opening of cyclic nucleotide-gated channels. Neuron 30, 689–698. Flynn, G. E., and Zagotta, W. N. (2003). A cysteine scan of the inner vestibule of cyclic nucleotide-gated channels reveals architecture and rearrangement of the pore. J. Gen. Physiol. 121, 563–582. Garneau, L., Klein, H., Banderali, U., Longpre-Lauzon, A., Parent, L., and Sauve, R. (2009). Hydrophobic interactions as key determinants to the KCa3.1 channel closed configuration: An analysis of KCa3.1 mutants constitutively active in zero Ca2þ. J. Biol. Chem. 284, 389–403. Grgic, I., Eichler, I., Heinau, P., Si, H., Brakemeier, S., Hoyer, J., and Kohler, R. (2005). Selective blockade of the intermediate-conductance Ca2þ-activated Kþ channel suppresses proliferation of microvascular and macrovascular endothelial cells and angiogenesis in vivo. Arterioscler. Thromb. Vasc. Biol. 25, 704–709. Grgic, I., Kiss, E., Kaistha, B. P., Busch, C., Kloss, M., Sautter, J., Muller, A., Kaistha, A., Schmidt, C., Raman, G., Wulff, H., Strutz, F., et al. (2009). Renal fibrosis is attenuated by targeted disruption of KCa3.1 potassium channels. Proc. Natl. Acad. Sci. USA 106, 14518–14523. Holmgren, M., Liu, Y., Xu, Y., and Yellen, G. (1996). On the use of thiol-modifying agents to determine channel topology. Neuropharmacology 35, 797–804. Jager, H., Dreker, T., Buck, A., Giehl, K., Gress, T., and Grissmer, S. (2004). Blockage of intermediate-conductance Ca2þ-activated Kþ channels inhibit human pancreatic cancer cell growth in vitro. Mol. Pharmacol. 65, 630–638. Karlin, A., and Akabas, M. H. (1998). Substituted-cysteine accessibility method. Methods Enzymol. 293, 123–145. Karplus, P. A. (1997). Hydrophobicity regained. Protein Sci. 6, 1302–1307. Khanna, R., Chang, M. C., Joiner, W. J., Kaczmarek, L. K., and Schlichter, L. C. (1999). hSK4/hIK1, a calmodulin-binding KCa channel in human T lymphocytes. Roles in proliferation and volume regulation. J. Biol. Chem. 274, 14838–14849. Klein, H., Garneau, L., Banderali, U., Simoes, M., Parent, L., and Sauve´, R. (2007). Structural determinants of the closed KCa3.1 channel pore in relation to channel gating: Results from a substituted cysteine accessibility analysis. J. Gen. Physiol. 129, 299–315. Klein, H., Garneau, L., Trinh, N. T., Prive, A., Dionne, F., Goupil, E., Thuringer, D., Parent, L., Brochiero, E., and Sauve, R. (2009). Inhibition of the KCa3.1 channels by AMP-activated protein kinase in human airway epithelial cells. Am. J. Physiol. Cell Physiol. 284, 389–403. Kohler, m., Hirschberg, B., Bond, C. T., Kinzie, J. M., Marrion, N. V., Maylie, J., and Adelman, J. P. (1996). Small-conductance, calcium-activated potassium channels from mammalian brain. Science 273, 1709–1714.

456

Line Garneau et al.

Kohler, R., Wulff, H., Eichler, I., Kneifel, M., Neumann, D., Knorr, A., Grgic, I., Kampfe, D., Si, H., Wibawa, J., Real, R., Borner, K., et al. (2003). Blockade of the intermediate-conductance calcium-activated potassium channel as a new therapeutic strategy for restenosis. Circulation 108, 1119–1125. Kuo, A., Gulbis, J. M., Antcliff, J. F., Rahman, T., Lowe, E. D., Zimmer, J., Cuthbertson, J., Ashcroft, F. M., Ezaki, T., and Doyle, D. A. (2003). Crystal structure of the potassium channel KirBac1.1 in the closed state. Science 300, 1922–1926. Laskowski, R. A., MacArthur, M. W., Moss, D. S., and Thornton, J. M. (1993). PROCHECK: A program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 26, 283–291. Long, S. B., Campbell, E. B., and MacKinnon, R. (2005). Crystal structure of a mammalian voltage-dependent Shaker family Kþ channel. Science 309, 897–903. Lu, Q., and Miller, C. (1995). Silver as a probe of pore-forming residues in a potassium channel. Science 268, 304–307. Marcus, Y. (1988). Ionic radii in aqueous solutions. Chem. Rev. 88, 1475–1498. Maylie, J., Bond, C. T., Herson, P. S., Lee, W. S., and Adelman, J. P. (2004). Small conductance Ca2þ-activated Kþ channels and calmodulin. J. Physiol. 554, 255–261. Ouadid-Ahidouch, H., Roudbaraki, M., Delcourt, P., Ahidouch, A., Joury, N., and Prevarskaya, N. (2004). Functional and molecular identification of intermediate-conductance Ca(2þ)-activated K(þ) channels in breast cancer cells: Association with cell cycle progression. Am. J. Physiol. Cell Physiol. 287, C125–C134. Rufo, P. A., Merlin, D., Riegler, M., Ferguson-Maltzman, M. H., Dickinson, B. L., Brugnara, C., Alper, S. L., and Lencer, W. I. (1997). The antifungal antibiotic, clotrimazole, inhibits chloride secretion by human intestinal T84 cells via blockade of distinct basolateral Kþ conductances. Demonstration of efficacy in intact rabbit colon and in an in vivo mouse model of cholera. J. Clin. Invest. 100, 3111–3120. Sali, A., and Blundell, T. L. (1993). Comparative protein modelling by satisfaction of spatial restraints. J. Mol. Biol. 234, 779–815. Schumacher, M. A., Rivard, A. F., Bachinger, H. P., and Adelman, J. P. (2001). Structure of the gating domain of a Ca2þ-activated Kþ channel complexed with Ca2þ/calmodulin. Nature 410, 1120–1124. Schumacher, M. A., Crum, M., and Miller, M. C. (2004). Crystal structures of apocalmodulin and an apocalmodulin/SK potassium channel gating domain complex. Structure (Camb.) 12, 849–860. Shannon, R. D. (1976). Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Crystallogr. A32, 751–767. Singh, S., Syme, C. A., Singh, A. K., Devor, D. C., Bridges, R. J., Lambert, L. C., DeLuca, A., and Frizzell, R. A. (2001). Benzimidazolone activators of chloride secretion: Potential therapeutics for cystic fibrosis and chronic obstructive pulmonary disease. Bicarbonate and chloride secretion in Calu-3 human airway epithelial cells. J. Gen. Physiol. 296, 600–611. Srivastava, S., Choudhury, P., Li, Z., Liu, G., Nadkarni, V., Ko, K., Coetzee, W. A., and Skolnik, E. Y. (2006a). Phosphatidylinositol 3-phosphate indirectly activates KCa3.1 via 14 amino acids in the carboxy terminus of KCa3.1. Mol. Biol. Cell 17, 146–154. Srivastava, S., Li, Z., Ko, K., Choudhury, P., Albaqumi, M., Johnson, A. K., Yan, Y., Backer, J. M., Unutmaz, D., Coetzee, W. A., and Skolnik, E. Y. (2006b). Histidine phosphorylation of the potassium channel KCa3.1 by nucleoside diphosphate kinase B is required for activation of KCa3.1 and CD4 T cells. Mol. Cell 24, 665–675. Szkotak, A. J., Murthy, M., MacVinish, L. J., Duszyk, M., and Cuthbert, A. W. (2004). 4Chloro-benzo[F]isoquinoline (CBIQ) activates CFTR chloride channels and KCNN4 potassium channels in Calu-3 human airway epithelial cells. Br. J. Pharmacol. 142, 531–542.

Constitutively Active KCa3.1

457

Webster, S. M., del Camino, D., Dekker, J. P., and Yellen, G. (2004). Intracellular gate opening in shaker Kþ channels defined by high-affinity metal bridges. Nature 428, 864–868. Wissmann, R., Bildl, W., Neumann, H., Rivard, A. F., Klocker, N., Weitz, D., Schulte, U., Adelman, J. P., Bentrop, D., and Fakler, B. (2002). A helical region in the C-terminus of small-conductance Ca2þ-activated Kþ channels controls assembly with apo-calmodulin. J. Biol. Chem. 277, 4558–4564. Wulff, H., Miller, M. J., Hansel, W., Grissmer, S., Cahalan, M. D., and Chandy, K. G. (2000). Design of a potent and selective inhibitor of the intermediate-conductance Ca2þ-activated Kþ channel, IKCa1: A potential immunosuppressant. Proc. Natl. Acad. Sci. USA 97, 8151–8156. Wulff, H., Knaus, H. G., Pennington, M., and Chandy, K. G. (2004). Kþ channel expression during B cell differentiation: Implications for immunomodulation and autoimmunity. J. Immunol. 173, 776–786. Xiao, J., Zhen, X. G., and Yang, J. (2003). Localization of PIP2 activation gate in inward rectifier Kþ channels. Nat. Neurosci. 6, 811–818. Yi, B. A., Lin, Y. F., Jan, Y. N., and Jan, L. Y. (2001). Yeast screen for constitutively active mutant G protein-activated potassium channels. Neuron 29, 657–667.