Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Activation of small conductance Ca²⁺-activated K⁺ channels (SK channels) by elevated intracellular Ca²⁺ underlies the slow afterhyperpolarization (sAHP) in neurons and regulates spike-firing frequency (Sah, 1996). Intracellular Ca²⁺ activates SK channels with submicromolar affinity (Kohler et al., 1996). Either Ca²⁺ influx through voltage-dependent Ca²⁺ channels or its release from intracellular stores is normally sufficient for SK channel activation (Sah, 1996). Cloning and mapping studies have identified three neuronal SK channel subunits, SK1, SK2, and SK3 (Kohler et al., 1996; Stocker et al., 1999). The SK2 subtype is expressed prominently in brain neurons and neuroendocrine glands; it likely constitutes the bulk of brain binding sites for the bee venom convulsant peptide apamin (Kohler et al., 1996; Stocker et al., 1999).

Chlorzoxazone (5-chloro-2-hydroxybenzoxazole; Parafon Forte DSC; see Fig. 2A below) is a centrally acting skeletal muscle relaxant (Elenbaas, 1980). The molecular and cellular mechanism of action of this drug is unknown, but it is structurally similar to 1-ethyl-2-benzimidazolinone (1-EBIO; see Fig. 2A). 1-EBIO enhances the activity of intermediate conductance Ca²⁺-activated K⁺ channels [IK channels (Devor et al., 1996b; Jensen et al., 1998)]. IK and SK channels belong to the same gene family and appear to share a common Ca²⁺-gating mechanism (Xia et al., 1998; Fanger et al., 1999). Zoxazolamine (2-amino-5-chlorobenzoxazole; see Fig. 2A) and NS 1619 (1,3-dihydro-1-[2-hydroxy-5-(trifluoromethyl)phenyl]-5-(trifluoromethyl)-2H-benzimidazol-2-one; see Fig. 5A) are also related structurally to chlorzoxazone and 1-EBIO. We therefore tested the effects of chlorzoxazone, 1-EBIO, zoxazolamine, and NS 1619 on rSK2 channels expressed in a mammalian cell line.

All four compounds affected rSK2 channel function. When applied externally chlorzoxazone, 1-EBIO, and zoxazolamine activated rSK2 channels in cells dialyzed with a nominally Ca²⁺-free intracellular solution. The responses to chlorzoxazone, 1-EBIO, and zoxazolamine declined at higher drug doses. Zoxazolamine, which partly inhibited the responses to chlorzoxazone and 1-EBIO, may be a partial agonist. In excised inside-out patches, intracellular 1-EBIO did not activate rSK2 channels when the intracellular free Ca²⁺ concentration was nominally zero. When the intracellular free Ca²⁺ was raised to 20 nM, 1-EBIO activated rSK2 channel currents in a concentration-dependent manner. NS 1619 did not activate rSK2 channels. Instead, this compound inhibited rSK2 channel currents activated by either chlorzoxazone, 1-EBIO, zoxazolamine, or elevated intracellular Ca²⁺.

Modulation of neuronal SK channels may be involved in the pharmacological and behavioral actions of these compounds in vivo and may provide novel pharmacological tools for investigating structure, function and modulation of native and recombinant neuronal SK channels.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Functional SK Channel Expression. All experiments were performed on recombinant rSK2 channels (Kohler et al., 1996) permanently expressed in a cell line derived from HEK293 cells. The line was generated by transfecting HEK293 cells with a plasmid encoding the rSK2 sequence and a geneticin-resistance gene contained within the mammalian expression vector pcDNA 3.1 (Invitrogen, Carlsbad, CA). Geneticin-resistant colonies were clonally purified, propagated, and tested for rSK2 channel expression by patch clamp. The cells were grown in minimal essential medium supplemented with 10% fetal calf serum, antibiotics (penicillin and streptomycin), sodium pyruvate, Glutamax, and 1 mg/ml geneticin (Life Technologies Inc., Rockville, MD). The cells were incubated at 37°C in a water-saturated 5% CO₂ atmosphere and passaged one to two times weekly. The cells were plated on 35-mm plastic Petri dishes (model 3001, Falcon, Becton Dickinson Co., Franklin Lakes, NJ) 1 to 2 days before experiments.

Electrophysiological Recording. Whole-cell patch clamp recording of rSK2 channel currents was performed as previously described (Dreixler et al., 2000a,b). Briefly, patch pipettes were pulled from thin-wall borosilicate glass, wax-coated, and fire-polished to 4- to 5-M tip resistance when filled with a nominally Ca²⁺-free intracellular solution composed of (in mM): 137.5 KMeSO₄, 1.0 MgCl₂, 10.0 K-BAPTA, 10.0 HEPES, 3.0 K₂ATP, 5.0 glucose; pH, 7.3. This solution was also applied to the intracellular aspect of inside-out excised macropatches in the experiment shown below in Fig. 4, either unmodified or modified to increase its free Ca²⁺ concentration to 20 nM. This was achieved by raising the total Ca²⁺ concentration to 0.91 mM. For the experiments shown in Figs. 1 and 5C (see below), we used an intracellular solution containing >1 µM free Ca²⁺ concentration; it was composed of (in mM): 137.5 KMeSO₄, 1.0 MgCl₂, 3.0 EGTA, 10.0 HEPES, 2.0 CaCl₂, 3.0 K₂ATP, 5.0 glucose; pH, 7.3. This free intracellular Ca²⁺ concentration is sufficient to maximally activate rSK2 channels (Kohler et al., 1996). The cells were voltage-clamped at 100-mV holding potential and constantly perfused with a modified mammalian Ringer's solution (30 mM K⁺ Ringer solution) composed of (in mM): 117.0 NaCl, 30.0 KCl, 2.0 CaCl₂, 1.0 MgCl₂, 10.0 HEPES, 5.0 glucose, 2.0 NaHCO₃; pH, 7.4. The calculated K⁺ reversal potential under these conditions was ~41 mV. Activation of rSK2 channels appeared as a sustained inward membrane current at 100 mV holding potential. Membrane current data were filtered at 33 Hz (8-pole Bessel) and digitized at 100 Hz using an EPC9 patch clamp amplifier (HEKA, Lambrecht, Germany). To determine the reversal potential, current-voltage (IV) relations were generated by ramping the membrane potential over the range 100 to +50 mV at 150 mV/s. Membrane currents were filtered at 1 kHz and digitized at 10 kHz. Five successive membrane current ramps were averaged. The average of five ramps, obtained under "background" conditions, was subtracted. The remaining current signal was plotted against the membrane voltage and least-square-fitted with a ninth order polynomial. The reversal potential was determined from this fit. When using antagonists (dequalinium, barium, and apamin in Fig. 1), background ramp currents were recorded at the peak of drug inhibition and subtracted from ramp currents recorded immediately before drug application. For the agonists in Fig. 2, the background ramp currents obtained immediately before drug application were subtracted from those obtained during the peak of the drug response. Inside-out macropatch current recordings were obtained using pipettes pulled from thick-walled borosilicate glass capillaries and fire-polished to have 4- to 6-M tip resistance when filled with an isotonic Ringer's solution composed of (in mM): 147 KCl; 2.0 CaCl₂; 1.0 MgCl₂; 5.0 glucose; 2.0 NaHCO₃; 10 HEPES (pH adjusted with KOH to 7.2). To maximally activate the rSK2 channels in the inside-out patches, the intracellular aspect of the patches was exposed to a solution containing 3.0 µM free Ca²⁺. It was composed of (in mM): 135 KMeSO₄; 10 HEPES; 10 HEDTA; 1.63 MgCl₂; and 3.6 CaCl₂ (pH adjusted with KOH to 7.2).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Expression of rSK2 channels in HEK293 cells dialyzed with high [Ca2+] intracellular solution. The traces in the top row (A) were recorded in HEK293 cells permanently expressing recombinant rSK2 channels; the traces in the middle row (B) were recorded in control, nontransfected HEK293 cells. Application of 10 µM dequalinium (first column), 10 mM barium (second column), and 1 nM apamin (third column) is indicated by the bar. The dashed line indicates the zero-holding current level. The time calibration was 20 s for the dequalinium- and barium-treated cells and 70 s for apamin-treated cells; the calibration of current amplitude was 1.5 nanoamperes (nA) for rSK2-expressed cells and 1 nA for control, nontransfected cells. The IV relations for dequalinium-, barium-, and apamin-sensitive currents, shown in C for both rSK2 channel-expressing cells and nontransfected cells (indicated by the arrowhead), were generated as described under Materials and Methods.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Chlorzoxazone and related compounds activate rSK2 channel currents in HEK293 cells. The top row (A) shows the chemical structure of chlorzoxazone, 1-EBIO, and zoxazolamine. The membrane current traces in the second row (B) were recorded with the application of chlorzoxazone (1 mM; left column), 1-EBIO (1 mM; middle column), and zoxazolamine (1 mM; right column); the traces in the third row (C) were recorded with the coapplication of 10 µM dequalinium with chlorzoxazone, 1-EBIO, and zoxazolamine. Holding potential and currents were 100 mV and ~70 to ~100 pA, respectively. Time and membrane current calibrations are 10 s and 0.5 nA, respectively. The bottom row (D) shows IV relations, generated as described under Materials and Methods, for the currents induced by chlorzoxazone, 1-EBIO, and zoxazolamine.

Chemicals. 1-EBIO was from Tocris Neuramin (Ballwin, MO); chlorzoxazone and zoxazolamine were from Aldrich (Milwaukee, WI); 50 mM dimethyl sulfoxide stocks of these compounds were used in this study. NS 1619 (Research Biochemicals, Natick, MA) was dissolved in dimethyl sulfoxide to make a 10 mM stock. Apamin (10 µM stock in 0.1% bovine serum albumin/water), dequalinium (10 mM in water), and all salts were form Sigma Chemical Co. or Fluka (St. Louis, MO), except KMeSO₄, which was from Pfaltz and Bauer (Waterbury, CT). All stocks were kept frozen and diluted directly into the 30 mM K⁺ Ringer solution before use.

Data Analysis. Membrane current responses to drugs were normalized to responses evoked by 1 mM 1-EBIO in the same cell, averaged (±S.E.M.), and plotted semilogarithmically against the dose. Each point is the average of six to nine independent determinations in different cells. The pooled, normalized dose-response data were least-square-fitted to the product of two logistic curves of the form:

(1)

where I is the current normalized to that induced by 1 mM 1-EBIO; C, the drug concentration; I_max, the maximum current; EC₅₀, and IC₅₀, the concentration for half-maximum activation and inhibition, respectively; and n1, and n2, the Hill number for activation and inhibition, respectively. For the curves in Fig. 3B, n2 was constrained to 2 to achieve least-square convergence. In the experiment shown in Fig. 4B, membrane patch currents were internally normalized to the response evoked by the 3 µM Ca²⁺-free intracellular solution. The data analysis was performed in Igor Pro (WaveMetrics, Lake Oswego, OR).

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Chlorzoxazone and related compounds activate rSK2 currents in a concentration-dependent manner. A, traces of rSK2 channel currents activated by increasing 1-EBIO concentrations in a cell voltage-clamped at 100 mV. The holding current before drug application was ~80 pA. Time and membrane current calibrations are 5 s and 0.4 nA, respectively. B, concentration-dependent induction of rSK2 currents by chlorzoxazone, 1-EBIO, and zoxazolamine. The currents were normalized to the response induced by 1 mM 1-EBIO and plotted as means ± S.E.M. (n = 6-9). The curve shows the least-square fitting to eq. 1.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   1-EBIO modulates rSK2 channel currents in excised inside-out macropatches. A, representative membrane current recordings from one patch. The intracellular aspect of the membrane patch was continuously perfused with nominally Ca2+-free intracellular solution. The bar indicates the duration of perfusion of test solutions. In the top trace the patch was exposed to 20 nM Ca2+. In the second trace perfusion with 20 nM Ca2+ with 0.3 mM 1-EBIO evoked an inward current due to rSK2 channel activation. In the third trace perfusion with 20 nM Ca2+ with 1 mM 1-EBIO evoked a larger current. In the bottom trace perfusion with 3.0 µM Ca2+ evokes the largest current due to maximal activation of rSK2 channels. Holding potential and current were 100 mV and ~42 pA, respectively. Time and membrane current calibrations are 10 s and 1.0 nA, respectively. B, pooled data summary. Membrane currents evoked by 20 nM Ca2+ alone or with 0.3 and 1.0 mM 1-EBIO were normalized to the response evoked by 3 µM Ca2+ in the same patch. Error bars represent mean ± S.E.M. (n = 5-6).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Functional Expression of rSK2 Channels in HEK293 Cells. We utilized an HEK293-derived cell line constitutively expressing rSK2 channels to study the effects of chlorzoxazone and related compounds. A high Ca²⁺ intracellular solution evoked large sustained inward currents when dialyzed into these cells voltage-clamped at 100 mV (Fig. 1A). These currents derived mostly from the activation of rSK2 channels (plus a small leak current component), because they were blocked by SK channel blockers apamin (1 nM), dequalinium (10 µM), and barium (10 mM; Fig. 1A). These currents were not observed in nontransfected HEK293 cells (Fig. 1, B and C) nor in transfected cells dialyzed with the nominally Ca²⁺-free solution (data not shown). The reversal potential of these currents was determined from ramp IV relations to be ~41 mV, identical to the calculated K⁺ reversal potential under our ionic conditions (Fig. 1C). Moreover, inside-out patches excised from the transfected cells, but not from control untransfected cells, expressed small conductance K⁺ channels activated by submicromolar Ca²⁺ concentrations (data not shown). Collectively, these results indicate that rSK2 channels mediate these currents.

Activation of rSK2 Channels by Chlorzoxazone, 1-EBIO, and Zoxazolamine. External application of these drugs to rSK2-expressing HEK293 cells dialyzed with a nominally Ca²⁺-free intracellular solution activated membrane currents (Fig. 2B) that were blocked by dequalinium (Fig. 2C) and reversed at the K⁺ reversal potential (~41 mV; Fig. 2D). We investigated whether 1-EBIO activated currents in nontransfected HEK293 cells. 1 mM 1-EBIO activated 1.2 ± 0.2 picoamperes (pA)/picofarad (pF) current in nontransfected HEK293 cells voltage-clamped at 100 mV (n = 7). This current increased ~100 fold to 115 ± 13 pA/pF in rSK2-channel expressing HEK293 cells (n = 14). The difference was highly significant (p < 0.001; unpaired Student's t test). Taken together, these results suggest that 1-EBIO and its structural relatives activate rSK2 channels.

View larger version (10K): [in this window] [in a new window]   Fig. 5.   Zoxazolamine occludes rSK2 channel currents activated by chlorzoxazone and 1-EBIO. As a control, the cell was first treated with chlorzoxazone or 1-EBIO (1 mM; first column), and zoxazolamine (1 mM; second column); then followed coapplication of 1 mM zoxazolamine with chlorzoxazone or 1-EBIO (third column). Following drug washout, chlorzoxazone or 1-EBIO were applied again (fourth column). Solid bars indicate drug treatment duration. The arrowheads indicate rebound current "bump" following drug washout. Recordings in the top and bottom rows were obtained in different cells. The holding potential was 100 mV; the holding current was ~180 pA in A and ~280 pA in B. Time calibration is 30 s; membrane current calibration is 2 nA in A and 1 nA in B.

Inhibition of rSK2 Channel Currents by NS 1619. NS 1619 (Fig. 6A) at 1, 3, or 10 µM, surprisingly did not activate rSK2 channels (Fig. 6B). Instead, this drug inhibited rSK2 channel currents activated by intracellular Ca²⁺ (Fig. 6C). Moreover, NS 1619 also inhibited the rSK2 channel currents activated by chlorzoxazone, 1-EBIO, and zoxazolamine. Figure 7 shows that 10 µM NS 1619 completely inhibited the response to 1 mM 1-EBIO, 1 mM zoxazolamine, and 1 mM chlorzoxazone (n = 5). These observations suggest that NS 1619 is an rSK2 channel antagonist.

View larger version (10K): [in this window] [in a new window]   Fig. 6.   NS 1619 inhibits rSK2 channel currents activated by intracellular Ca2+. A, NS 1619 structure. B, membrane current trace recorded in an rSK2-expressing HEK293 cell dialyzed with a nominally Ca2+-free intracellular solution. C, membrane current trace recorded in an rSK2-expressing HEK293 cell dialyzed with high Ca2+ intracellular solution. The bar indicates the application of 10 µM NS 1619. The holding potential was 100 mV; the dashed line indicates zero holding current. The time calibration is 10 s for B and 20 s for C; the current amplitude calibration is 1 nA.

View larger version (9K): [in this window] [in a new window]   Fig. 7.   NS 1619 inhibits rSK2 channel currents activated by chlorzoxazone and related compounds. Drug concentrations were 10 µM for NS 1619, and 1 mM for chlorzoxazone, 1-EBIO, and zoxazolamine. Membrane current traces were recorded in rSK2-expressing HEK293 cells dialyzed with a nominally Ca2+-free intracellular solution. Cells were first treated with chlorzoxazone, 1-EBIO, and zoxazolamine (left column). After drug washout, the cells were pre-equilibrated with NS 1619 for 60 s. NS 1619 pretreatment did not alter the membrane current. Chlorzoxazone, 1-EBIO, and zoxazolamine were then applied in the continued presence of NS 1619 (middle column). Following drug washout, chlorzoxazone, 1-EBIO, and zoxazolamine were applied again (right column). Solid bars denote agonist application duration. Holding potential and membrane current were 100 mV and ~210 pA, respectively. Time and membrane current amplitude calibrations are 30 s and 2 nA, respectively.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We report six novel findings. First, chlorzoxazone and the structurally related compounds 1-EBIO and zoxazolamine activate recombinant rSK2 channels when applied extracellularly. Second, the extent of activation depends on the chemical structure and the concentration. Third, these compounds activate rSK2 channels in the nominal absence of intracellular Ca²⁺ in whole-cell experiments; in excised patch experiments a minimum amount of Ca²⁺, ~20 nM, is required for rSK2 channel activation by these drugs. Fourth, at high concentrations, some of these compounds simultaneously activate and block rSK2 channels. Fifth, when two compounds were coapplied, their effect was not additive, suggesting that they may share a common mode of action. Sixth, NS 1619, another structurally related compound, did not activate but inhibited rSK2 channel currents.

Therapeutic plasma concentrations of chlorzoxazone can reach ~180 µM (Desiraju et al., 1983). Our results with rSK2 channels suggest that therapeutic concentrations of chlorzoxazone are likely to substantially activate SK channels in vivo. This activation would be consistent with the muscle-relaxant effects of chlorzoxazone. Our results confirm and extend a recent study showing that chlorzoxazone activates the structurally related IK channels encoded by the SK4 gene (Singh et al., 2000).

We observed that 1-EBIO activates rSK2 channels. 1-EBIO has previously been shown to activate native and recombinant IK channels (Devor et al., 1996a; Jensen et al., 1998). 1-EBIO action on IK channels is consistent with a leftward shift of the Ca²⁺ activation curve (Pedersen et al., 1999). 1-EBIO was reported to require a nonzero intracellular Ca²⁺ concentration to activate IK channels (Devor et al., 1996a; Pedersen et al., 1999). In contrast, we observed that extracellular 1-EBIO activates whole-cell rSK2 channel currents in the nominal absence of intracellular Ca²⁺. Our intracellular solution contained 10 mM BAPTA and no added Ca²⁺. We have previously determined that the total Ca²⁺ contamination of our solutions was ~1.2 µM (Cao and Houamed, 1999). Making a conservative assumption that our total Ca²⁺ contamination is 10 µM we calculate that 10 mM BAPTA would buffer the free Ca²⁺ concentration in our intracellular solution to <0.3 nM. The minimum free Ca²⁺ concentration required for IK channel activation with 1-EBIO was ~30 nM (Pedersen et al., 1999). Since rSK2 channels are more sensitive to Ca²⁺ than IK channels (Ishii et al., 1997) it is possible that some Ca²⁺ must be present for 1-EBIO to activate rSK2 channels, but the amount is in the subnanomolar range. Moreover, in this study we used higher 1-EBIO concentrations on rSK2 channels than others have used on native and recombinant IK channels (Devor et al., 1996a; Pedersen et al., 1999). It is also possible that low 1-EBIO concentrations allosterically modify Ca²⁺-dependent gating and high concentrations open the channels directly, rather like the action of barbiturates on -aminobutyric acid A receptor channels (Akaike et al., 1985).

Our excised patch experiment, shown in Fig. 4, is in agreement with previous work suggesting that a minimum amount of free Ca²⁺ must be present for SK channels to be activated by chlorzoxazone and its analogs (Singh et al., 2000). There are two possible explanations for the apparent discrepancy between our whole-cell and excised patch observations. First, it is possible that our intracellular dialysis with the whole-cell patch pipette was incomplete and did not lower the intracellular Ca²⁺ concentration to subnanomolar levels as we anticipated. We believe this is unlikely. Following whole-cell break-in, the intracellular milieu equilibrates with the patch pipette contents with a monoexponential time course (Pusch and Neher, 1988). Pipette access resistance, cell size, and diffusion coefficients determine the equilibration time constant (Pusch and Neher, 1988). Our experiments were done on single isolated HEK293 cells. These cells are small and geometrically simple, suggesting that they would be readily dialyzed. We initiated drug testing no earlier than 6 min after whole-cell break-in, and continued for up to 30 min thereafter. During this time the drug responses were stable and reproducible, suggesting that the intracellular ionic milieu had already equilibrated with the pipette solution by the beginning of the recording. Furthermore, when the whole-cell pipette solution contained a high free Ca²⁺ concentration (Figs. 1A and 7) the SK channel current was activated and reached steady state in less than 3 min, confirming rapid intracellular dialysis.

The second possible explanation for the observed differences in whole-cell and excised-patch experiments may reflect the mode of action of the drugs. In the whole-cell experiments the drugs were applied externally, whereas in the excised-patch experiment they were applied internally. Additionally, endogenous cellular factors may influence drug efficacy. Such factors may include intracellular enzymes or cytoskeletal elements. Further studies on the mechanism and site of action of these drugs may resolve this anomaly.

NS 1619 activates large conductance Ca²⁺-activated K⁺ channels (BK channels) (Olesen et al., 1994). Our surprising finding, that it did not activate, but inhibited, rSK2 channels, is consistent with recent studies showing opposite actions of NS 1619 and 1-EBIO on IK channels (Pena and Rane, 1999).

In conclusion, we have shown that chlorzoxazone, 1-EBIO, and zoxazolamine activate recombinant rat brain SK2 channels in a Ca²⁺-independent manner. The effects of chlorzoxazone on SK2 channels are consistent with the muscle-relaxant effects of this drug. This novel mode of action should render these drugs useful tools for studying the pharmacology and physiology of recombinant and native neuronal SK channels. The recent advent of transgenic animals lacking specific SK channel subtypes (Bond et al., 2000) should allow further study on the interaction of chlorzoxazone and its structural analogs with SK channels at the organism level.