Sodium-activated potassium (KNa) channels have been suggested to set the resting potential, to modulate slow after-hyperpolarizations, and to control bursting behavior or spike frequency adaptation (Trends Neurosci 28:422–428, 2005). One of the genes that encodes KNa channels is called Slack (Kcnt1, Slo2.2). Studies found that Slack channels were highly expressed in nociceptive dorsal root ganglion neurons and modulated their firing frequency (J Neurosci 30:14165–14172, 2010). Therefore, Slack channel openers are of significant interest as putative analgesic drugs. We screened the library of pharmacologically active compounds with recombinant human Slack channels expressed in Chinese hamster ovary cells, by using rubidium efflux measurements with atomic absorption spectrometry. Riluzole at 500 μM was used as a reference agonist. The antipsychotic drug loxapine and the anthelmintic drug niclosamide were both found to activate Slack channels, which was confirmed by using manual patch-clamp analyses (EC50 = 4.4 μM and EC50 = 2.9 μM, respectively). Psychotropic drugs structurally related to loxapine were also evaluated in patch-clamp experiments, but none was found to be as active as loxapine. Loxapine properties were confirmed at the single-channel level with recombinant rat Slack channels. In dorsal root ganglion neurons, loxapine was found to behave as an opener of native KNa channels and to increase the rheobase of action potential. This study identifies new KNa channel pharmacological tools, which will be useful for further Slack channel investigations.
Slack channels belong to the Ca2+-activated K+ channel family because they contain tandem regulator of K+ conductance (RCK) domains (Jiang et al., 2002). However, Slack channels (Slo2.2) and the closely related Slick channels (Slo2.1) are activated by Na+ and Cl− and not Ca2+. Mutational analyses indicated that Na+ binding and gating occur in RCK domain 2 (Zhang et al., 2010). Slack channels are widely expressed throughout the rat central nervous system, including the substantia nigra (SN), but also are abundantly expressed in peripheral dorsal root ganglion (DRG) neurons (Bhattacharjee et al., 2002; Tamsett et al., 2009). Because Slack channels are activated by both voltage and cytosolic factors such as intracellular Na+ and Cl−, they are ideal effectors of negative feedback during neuronal excitation. It was shown in lamprey spinal cord neurons that AMPA receptor activation was linked, via Na+ transients, to an increase in KNa current amplitude (Nanou et al., 2008). Likewise, tetrodotoxin was shown to decrease strongly the outward K+ current evoked by depolarization in rat olfactory neurons, and a tetrodotoxin-sensitive K+ current was eliminated by a Slack small interfering RNA (Budelli et al., 2009). In locus ceruleus, glutamate-induced postactivation inhibition was found to be mediated by AMPA/kainate receptors and bithionol-sensitive KNa currents (Zamalloa et al., 2009). These findings suggest that Slack activation may limit abnormal neuronal activity in pathological conditions such as epilepsy or pain. With the exception of bithionol, there are few pharmacological tools available for study of the functional role of KNa channels in neurons, particularly in a pathological context. To evaluate the therapeutic potential of KNa channel openers, we sought to identify novel lead compounds that activate Slack channels.
The screening of ion channels at medium throughput has been facilitated by the availability of automated patch-clamp devices. However, their use remains rather expensive for medium-sized libraries (i.e., more than 10,000 compounds) and is not fully appropriate for assessment of compounds or channels that can cause background currents, which may be merged with leakage currents. Therefore, we have designed a functional assay assessing Rb+ efflux, through atomic absorption spectrometry (AAS), in Chinese hamster ovary (CHO) cells stably expressing recombinant human Slack channels (CHO-hSlack cells). This technique has already been shown to be rather efficient and of low cost during K+ channel screening (Terstappen, 2004). Riluzole, which was found previously to be a Slack channel opener (our unpublished data), was used as a reference compound, and the LOPAC 1280 was screened as a test validation.
By using this screening approach, we identified loxapine and niclosamide as novel Slack channel openers. Patch-clamp experiments with both recombinant Slack and native KNa channels were used to confirm these findings.
Materials and Methods
Riluzole, loxapine, niclosamide, bithionol, and amoxapine were purchased from Sigma-Aldrich (Saint Quentin Fallavier, France). Quetiapine, clozapine, and olanzapine were purchased from Interchim (Montlucon, France). Drugs were diluted as 25 mM stock solutions in dimethyl sulfoxide (DMSO), the concentration of which did not exceed 0.08% in the final solutions used.
Atomic Absorption Spectrometry Experiments.
The reference compound riluzole hydrochloride was purchased from Tocris Bioscience (Bristol, UK) and was diluted to 100 mM in DMSO, the concentration of which did not exceed 1% in the final solutions used. The LOPAC was purchased from Sigma-Aldrich.
DRG neurons were isolated and cultured as described previously (Nuwer et al., 2010). In brief, DRGs were dissected from embryonic day 15 embryos of Sprague-Dawley rats. The ganglia were dissociated with 2.5 mg/ml trypsin (Invitrogen, Paisley, UK) for 40 min. Neurons were plated on coverslips coated with poly-d-lysine (Sigma-Aldrich) and laminin (Invitrogen) and were maintained in serum-free medium containing 100 ng/ml nerve growth factor (Harlan Bioproducts for Science, Indianapolis, IN). One day after dissection, neurons were treated for 2 days with 1 μM cytosine-d-arabinofuranoside (Sigma-Aldrich), an inhibitor of DNA synthesis, to inhibit the growth of non-neuronal cells. Cells were allowed to recover for 2 days before manipulation. All experimental protocols were approved by the University at Buffalo institutional animal care and use committee and conformed to National Institutes of Health guidelines.
A stable CHO cell line expressing the human Slack channel (CHO-hSlack) was generated by the Lead Identification Technologies Department of Sanofi-Aventis Research and Development (Frankfurt, Germany) from the parental Flp-In-T-Rex CHO host cell line. Cells were cotransfected with the human KCNT1 gene (GenBank accession number NM_020822) inserted into the pCDNA5-FRT-TO plasmid and with the pOG44 plasmid containing the Flp recombinase. By using a similar protocol, a CHO cell line stably expressing human MaxiK channels (encoded by the KCNMA1 gene; GenBank accession number NM_002247) together with the KCNMB4 subunit (GenBank accession number NM_014505.1) was also produced by using an internal ribosome-entry site structure (CHO-MaxiK).
CHO-hSlack, CHO-MaxiK, and parental Flp-In-T-Rex cells were grown in culture medium composed of Dulbecco's modified Eagle's medium-F12 medium (Invitrogen) supplemented with 10% fetal bovine serum and 30 μg/ml blasticidin, 300 μg/ml hygromycin, or 100 μg/ml phleomycin D1, respectively (Invitrogen), and were maintained in an atmosphere of 95% air/5% CO2 at 37°C. They were dissociated for passage by using cell dissociation buffer (enzyme-free, phosphate-buffered saline-based; Invitrogen).
For single-channel studies, we used human embryonic kidney cells (HEK 293 cells) stably expressing rat Slack channels (Yang et al., 2006), from Yale University (New Haven, CT). Cells were cultured on 35-mm dishes in modified low-sodium Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and penicillin-streptomycin (Invitrogen) and were maintained in a 5% CO2 incubator at 37°C.
Atomic Absorption Spectrometry Studies
Compounds and reagents were diluted in Hanks' balanced salt solution (HBSS) that contained 5.4 mM KCl, 0.44 mM KH2PO4, 4.17 mM NaHCO3, 137.9 mM NaCl, 0.34 mM Na2HPO4, 5.5 mM glucose, and 2 mM HEPES (pH 7.4). The day before the experiment, CHO-Slack cells were seeded at a density of 2 × 104 cells per well in poly-d-lysine-coated, 384-well plates and were incubated for 24 h in culture medium containing 20 mM rubidium and 2 μg/ml tetracycline (Sigma-Aldrich) (final volume, 40 μl per well). The day of the experiment, the cells were washed three times with HBSS and then incubated for 20 min at room temperature in HBSS containing 30 μM concentrations of the tested compounds (final volume, 100 μl). Test compounds were prepared in 100% DMSO and diluted to the desired concentration in HBSS. After 20 min, the supernatant was pipetted and cells were lysed with 1% Triton X-100. The rubidium contents of the cell supernatant ([Rb+]sup) and lysate ([Rb+]lys) were determined by using an ICR 8000 atomic absorption spectrometer (Aurora Biomed, Vancouver, Canada) (Gill et al., 2003; Parihar et al., 2003). Calibration of the instrument was performed before each experiment, by using increasing Rb+ concentrations.
Whole-Cell Patch-Clamp Experiments with CHO-hSlack Cell Line.
For whole-cell patch-clamp experiments, CHO-hSlack or CHO-MaxiK cells were seeded on glass coverslips and were incubated for 20 to 24 h in cultivation medium containing a 1 μg/ml concentration of the inducing agent doxycycline (BD Biosciences, San Jose, CA). Experimental chambers (RC-26-GLP; Warner Instruments, Hamden, CT) containing seeded coverslips were placed on the stage of an inverted microscope (IMT2; Olympus France, Rungis, France) equipped with Hoffman optics (Modulation Contrast, New York, NY), and the cells were viewed at a total magnification of 400×. A gravity-fed perfusion valve-control system was used (VC-66CST; Warner Instruments), driven by Clampex (version 9.2; Molecular Devices, Sunnyvale, CA) and connected to a 12-way manifold ended by a glass tube (500-μm opening) placed less than 1 mm from the recorded cells. Pipettes were pulled from thick-walled borosilicate glass capillaries (Harvard Apparatus, Edenbridge, UK) on a horizontal two-stage puller (P97; Sutter Instruments, Novato, CA) and had a resistance of 5 to 10 MΩ when filled with the pipette solution (see below). Pipettes were brought into contact with the cells with a three-dimensional piezoelectric micromanipulator (Sutter MP 285; Dipsi Industrie, Chatillon, France). Whole-cell currents were recorded with a MultiClamp 700B amplifier (Molecular Devices) driven by MultiClamp 2.1 software (Molecular Devices).
The standard extracellular solution contained 147 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, and 10 mM HEPES; the pH was adjusted to 7.4 with 1 M NaOH. For Slack current recordings, the pipette was filled with an intracellular pipette medium containing 110 mM potassium aspartate, 5 mM EGTA, 10 mM KCl, 10 mM NaCl, 10 mM HEPES, 1 mM MgCl2, and 1 mM CaCl2; the pH was adjusted to 7.2 with 1 M KOH. For MaxiK current recordings, the composition of the intracellular pipette medium was 145 mM KCl, 2 mM MgCl2, 10 mM EGTA, 5 mM ATP, and 10 mM HEPES; the pH was set 7.4 with KOH.
Excised-Patch Channel Recordings.
For inside-out patch-clamp recordings, the pipette solution contained 10 mM NaCl, 130 mM KCl, 10 mM HEPES, 5 mM EGTA, 1 mM MgCl2, and 1 mM tetraethylammonium chloride. The bath solution contained 130 mM KCl, 10 mM NaCl, 10 mM HEPES, and 5 mM EGTA. The pH of all solutions was adjusted to 7.3 with KOH. Inside-out patches were perfused with the SmartSquirt small-volume delivery system (Automate Scientific, Berkley, CA), by using a 100-μm perfusion tip with a flow rate of 0.01 ml/min, and varying concentrations of loxapine were used in the perfusion.
Action Potential Recordings with DRG Neurons.
All experiments were performed at room temperature. Whole-cell, current-clamp recordings were made with a MultiClamp 700B amplifier, stored digitally with an Axon DigiData interface (1322 series), and analyzed off-line with ClampFit software (Molecular Devices). Electrode impedance was 3 to 5 MΩ when electrodes were filled with saline solution containing 124 mM potassium gluconate, 2 mM MgCl2, 13.2 mM NaCl, 1 mM EGTA, 10 mM HEPES, 4 mM Mg-ATP, and 0.3 mM Na-GTP (pH 7.2). The bath solution contained 140 mM NaCl, 5.4 mM KCl, 1 mM CaCl2, 1 mM MgCl2, 15.6 mM HEPES, and 10 mM glucose (pH 7.4). Neurons were accepted for study only when they exhibited a resting membrane potential more negative than −40 mV and an input resistance of more than 100 MΩ after establishment of the whole-cell mode. Depolarizing current steps in increments of 10 pA from −10 to 200 pA (20-ms duration) were used to measure the rheobase.
Atomic Absorption Spectrometry Studies.
The Rb+ efflux ratio was calculated as follows: Rb+ efflux ratio = [Rb+]sup/([Rb+]sup + [Rb+]lys). Dose-response experiments corresponding to eight concentrations were performed in quadruplicate. The effects of compounds were expressed as percentages of baseline values (effect of 1% DMSO). Dose-response curves and EC50 values were generated by using a standard, four-parameter, logistic, nonlinear, regression analysis, with GraphPad Prism 4.0 dedicated software (GraphPad Software Inc., San Diego, CA).
For whole-cell current recordings, all statistical analyses were performed on fold values after logarithmic transformation, to improve the heterogeneity of variance. The geometric mean and the 95% confidence interval were used as descriptive statistics. Depending on the objective of the study, Student's t test was performed to show a difference of treatment, compared with its own control. One-way analyses of variance were performed to show a difference between compounds or concentrations (Newmann-Keuls test) or equivalence, compared with a reference compound. EC50 estimation was performed through nonlinear regression with SAS 9.1 (SAS Institute Inc., Cary, NC), by using a four-parameter logistic model with fold values. For single-channel analyses, pClamp 9.0 software (Molecular Devices) was used. Data derived from these experiments were expressed as mean ± S.E. Statistically significant differences were assessed by using Student's t test.
Identification of Slack Channel Openers by Using Rb+ Efflux Measurements
For these experiments, riluzole (500 μM) was used as the reference agonist. This compound was previously found serendipitously to be an effective Slack opener (our unpublished data). The signal/noise ratio ranged between 2 and 2.5. Of 1280 compounds of the LOPAC, 42 compounds were identified as putative Slack channel openers (hit ratio, 3%) (Fig. 1, A and B). After confirmation of the compounds' lack of activity with the parental CHO cell line, their EC50 values were determined. Among these compounds were the antipsychotic drug loxapine (EC50 = 3.5 μM) and the anthelmintic drug niclosamide (EC50 = 0.7 μM). For comparison, riluzole increased Rb+ efflux with an EC50 value of 97 μM (Fig. 1C).
Whole-Cell Recordings with CHO-hSlack Cell Line.
With the use of a voltage-step protocol (from −90 to −20 mV for 200 ms), 10 μM loxapine, 10 μM niclosamide, and 100 μM riluzole were found to increase Slack current amplitude significantly, by 10.7-fold (95% confidence interval, 7.1–16.1-fold; p < 0.0001), 3.9-fold (95% confidence interval, 2.1–7.1-fold; p = 0.0015), and 6.5-fold (95% confidence interval, 3.4–12.4-fold; p = 0.0012), respectively, when measured at the end of the depolarizing voltage step (Fig. 2, A and B). At 10 μM, loxapine was found to be significantly more efficient than niclosamide (p = 0.0069). In the presence of loxapine and niclosamide, Slack current activation kinetics seemed to be accelerated; in the presence of riluzole, however, the typical slow activation of Slack current was observed. Loxapine, niclosamide, and riluzole obviously induced increases in steady-state inward current amplitude, measured at the holding voltage.
With a 2-s ramp protocol, depolarizing CHO-hSlack cell membranes from −120 to +40 mV, loxapine and niclosamide at 10 μM were found to increase Slack current amplitudes in both the inward and outward directions. In the presence of these two compounds, the current-voltage relationship became almost linear, even in the inward direction. This observation suggests that these compounds were able to turn Slack channels into a type of background potassium channels that are poorly sensitive to voltage (see data below on loxapine- and niclosamide-induced steady-state currents at −100 mV). For outward currents measured at +40 mV, loxapine and niclosamide increased the amplitude by 4.8-fold (95% confidence interval, 3.3–6.8-fold; p < 0.0001) and 3.2-fold (95% confidence interval, 2.1–5-fold; p = 0.0001), respectively. Riluzole was tested at 100 μM and increased Slack current in the outward direction by 3.7-fold (95% confidence interval, 1.3–10.2-fold; p = 0.0242) at +40 mV, being less effective at increasing inward currents (Fig. 2, C and D). With this ramp protocol, loxapine was found to be the most efficient of the three compounds.
As described above, loxapine and niclosamide were found to induce reproducibly inward Slack currents at potentials more negative than the K+ equilibrium potential. We used this effect to compare the two compounds and to build the concentration-response curves. At a holding voltage of −100 mV, loxapine induced strong inward currents in a concentration-dependent manner, with a calculated EC50 value of 4.4 μM (95% confidence interval, 2.7–7.2 μM) and a maximal increase in current amplitude of 8.9-fold (95% confidence interval, 8.1–9.8-fold) (Fig. 3, A and C). Even at the low concentration of 0.3 μM, loxapine was found to increase Slack steady-state currents significantly (p = 0.001). Niclosamide was found to be of similar potency, with an EC50 value of 2.9 μM (95% confidence interval, 1.3–6.5 μM), but less efficient, with a maximal current increase of 3.0-fold (95% confidence interval, 2.2–4.1-fold) (Fig. 3, B and D). As shown in Fig. 3B, at high concentrations (10 and 30 μM) the niclosamide effect had a propensity to decrease rapidly with time, which suggests some kind of desensitizing or open channel-blocking effect that was not observed with loxapine. In confirmation of the results described above, loxapine was found to be significantly more efficient (∼2-fold) than niclosamide at every concentration tested (p < 0.001 at 3 and 10 μM, p < 0.01 at 0.3 and 1 μM, and p < 0.05 at 30 μM).
To evaluate whether the Slack channel-opening properties of loxapine were linked to its typical tricyclic structure, the effects of structurally related antipsychotic or antidepressant drugs on steady-state Slack current were evaluated at 10 μM (olanzapine, quetiapine, clozapine, and amoxapine). As shown in Fig. 4, amoxapine and to a lesser extent quetiapine were the most efficient, inducing increases in Slack steady-state current amplitude of 3- and 2.2-fold, respectively. Clozapine and olanzapine were almost ineffective, which suggests that Slack channel activation is not a general property of tricyclic antipsychotic drugs.
Whole-Cell Recordings with CHO-MaxiK Cell Line.
Bithionol was shown previously to be an opener of MaxiK channels (Yang et al., 2006). Therefore, it was of interest to compare loxapine and bithionol effects on MaxiK currents, evoked by depolarizing CHO-MaxiK cells from a holding voltage of −85 mV to 20 mV over 100 ms (F = 0.1 Hz). Both compounds were superfused at 10 μM. As expected, bithionol behaved as a powerful reversible MaxiK channel opener, increasing current amplitude by 11.5-fold (95% confidence interval, 6.3–21-fold; p < 0.0001). In contrast, loxapine was inactive (1.05-fold; 95% confidence interval, 0.96–1.15-fold) (Fig. 5, A and B).
Effects of Loxapine on Recombinant Rat Slack Channels and Native KNa Channels.
We evaluated the effects of loxapine directly on recombinant rat Slack channels by conducting excised inside-out patch recordings from a stable rat Slack channel-expressing cell line (Yang et al., 2006). We found that direct application of loxapine to patches clearly caused an increase in Slack channel activity, which was reversed after washing (n = 11; p < 0.01) (Fig. 6, A–C). We also found a time-dependent increase in channel activity during outside-out patch recordings (n = 4; p < 0.05 at 3 min) (Fig. 6D). The time-dependent effects suggest that loxapine does not act at the pore of the channel but likely acts at the RCK domains internally and affects the gating of Slack. We also tested loxapine on native KNa channels recorded from cultured DRG neurons (Fig. 7, A and B). Embryonic DRG neurons express both Slack (Nuwer et al., 2010) and Slick (A. Bhattacharjee, unpublished observations) channel subunits. In excised patch recordings, native KNa channels were activated by loxapine in a reversible manner (n = 4).
Effect of Loxapine on Neuronal Excitability.
We tested the threshold of action potential generation in neurons incubated with 10 μM loxapine. We found that loxapine caused a significant increase in rheobase (n = 13; p < 0.05), compared with neurons treated with DMSO only (Fig. 7, C and D). In addition, the resting potential of loxapine-treated neurons was −57.9 ± 2.8 mV, compared with −51.9 ± 3 mV for control neurons (p < 0.05). These data support the specific effects of loxapine on native KNa channels and suggest that loxapine can directly affect the excitability of neurons.
To our knowledge, this is the first report of the development of a screening assay, in a medium-throughput format, targeting Slack channels. This assay, based on AAS, was performed with physiological intracellular Na+ concentrations. It has been well demonstrated for several decades that radioactive Rb+ can be used as an analog for K+ in flux-based assays. However, the use of radioactivity is a limitation for high-capacity screening. More recently, AAS methods have been developed for measurement of nonradioactive Rb+ flux through several K+ channels, including Kv1.3 (Gill et al., 2007), KCNQ2/3 (Scott et al., 2003), MaxiK (Parihar et al., 2003), and hERG (Chaudhary et al., 2006) channels. These results confirm that Rb+ efflux measurement with AAS represents a reliable, safe, low-cost alternative for K+ channel-screening purposes (Parihar et al., 2003; Terstappen, 2004; Gill et al., 2007). EC50 values determined by using AAS and patch-clamp methods were in good keeping for loxapine (3.5 and 4.4 μM, respectively) but showed some discrepancies for niclosamide (0.7 and 2.9 μM, respectively). Similar findings were described for KCNQ2 openers, with potencies varying from 2- to 10-fold and with greater affinities being found with AAS, compared with patch-clamp data (Wang et al., 2004).
This screening campaign using the LOPAC has led to the identification of loxapine (a first-generation antipsychotic drug that is still being prescribed) as an efficient opener of Slack channels. This new property has been confirmed with human and rat recombinant Slack channels and native KNa channels in DRG neurons. Slack channel studies have been hampered by the lack of efficient selective pharmacological tools; to date, only bithionol (an anthelmintic veterinary drug) has been used as a reference Slack channel opener. However, this compound was reported also to be an activator of MaxiK channels (Yang et al., 2006). This property has not been found for loxapine, which should facilitate the functional study of native Slack channels in preparations that express both Slack and MaxiK channels. Among the two other Slack channel openers that we have identified (i.e., niclosamide and riluzole), it is noteworthy that niclosamide is another anthelmintic drug that is still on the market. Slack-like channels are well expressed in nematodes such as Caenorhabditis elegans, and this species is classically used for anthelmintic drug screening (Wei et al., 1996; Kaewintajuk et al., 2010). The possibility cannot be ruled out that the Slack channel-opening properties of bithionol and niclosamide might be part of their antiparasitic properties. This could be addressed by studying the toxicity of these 2 molecules in C. elegans mutants that were reported previously to have strongly decreased Slo2-mediated currents (Yuan et al., 2003). Clinically, the possible effects of niclosamide or bithionol on central nervous system excitability via Slack channel activation will not be relevant because these compounds are poorly absorbed and do not cross the blood-brain barrier (Dagorn, 1982; Bagheri et al., 2004).
In addition, we are the first to demonstrate that riluzole, a compound prescribed for patients with amyotrophic lateral sclerosis, opens Slack channels in a concentration range similar to that found to be active at TRAAK or TREK-1 channels (Duprat et al., 2000). One study showed that riluzole at 20 μM was able to inhibit KNa currents attributed to Slack channels in cultured olfactory neurons (Budelli et al., 2009). As suggested by the authors, riluzole's effects can probably be attributed to blockade of the late component of Na+ currents. These results are not in opposition to those we report, which were obtained with higher riluzole concentrations. Given the rather high estimated EC50 value we found (97 μM in AAS studies; not assessed with patch-clamp experiments), it is unlikely that this new property of riluzole is involved in its therapeutic properties. However, it should be taken into account for in vitro or in vivo pharmacological studies performed with high riluzole concentrations, such as those needed to activate TRAAK or TREK-1 channels (Duprat et al., 2000). As an example, a study showed that riluzole (100–500 μM) applied locally in vivo to injured DRG neurons in a rat model of allodynia decreased the spontaneous firing rate of A-fibers. This effect was attributed to the inhibition of noninactivating Na+ currents, but an increase in KNa currents at such high concentrations cannot be ruled out (Xie et al., 2011).
In attempts to correlate in vitro data with plasma concentrations achieved in patients and clinical effects, loxapine seems to be a rather complex drug. The steady-state plasma levels of loxapine vary (∼9-fold range reported) among individuals treated chronically with the same oral dosage (Cooper et al., 1979). Furthermore, loxapine is extensively metabolized by cytochrome P450 to amoxapine, 7-hydroxyloxapine, and 8-hydroxyloxapine, which are further metabolized to 7-hydroxyamoxapine and 8-hydroxyamoxapine (Heel et al., 1978; Wong et al., 2012). In the very recent study by Wong et al. (2012), which was conducted with rats, it was shown that, 4 h after administration of 1 mg of loxapine, amoxapine, 7-hydroxyamoxapine, and 7-hydroxyloxapine plasma concentrations were approximately 15 to 20 times higher than that of loxapine. 7-Hydroxyloxapine was found to accumulate in the brain, achieving concentrations as high as 124 ng/g in the striatum, whereas loxapine, amoxapine, and 7-hydroxyamoxapine concentrations were less than 5 ng/g. Despite interspecies variability in loxapine metabolism (Bun et al., 2003), the antipsychotic effects of this drug in humans are more correlated with the plasma levels of its hydroxylated and hydroxylated desmethyl metabolites, rather than the very low parent drug concentrations (Simpson et al., 1978). The maximal plasma concentration after oral administration of 50 mg of loxapine is approximately 30 ng/ml (0.09 μM). This is 2 orders of magnitude lower than the EC50 value we calculated (EC50 = 4.4 μM). Significant increases in Slack current amplitude were observed for concentrations as low as 0.3 μM, however, and maximal concentration values as high as 0.4 μM (135 μg/l) were noted shortly after nasal administration of 10 mg of loxapine (Spyker et al., 2010). Together, these data strongly suggest that direct Slack channel activation by the unmetabolized form of loxapine during chronic treatment is clinically unlikely. However, it would be of great value to evaluate the Slack channel-modulating properties of loxapine and amoxapine hydroxylated derivatives. With respect to side effects, loxapine is known to be among the tricyclic antipsychotic agents most prone to inducing extrapyramidal syndrome (EPS), as shown by the very high rate of antiparkinsonism drug coprescription associated with loxapine use (Yang et al., 2007). Moreover, the study conducted by Yang et al. (2007) confirmed previously published studies, supporting a quantitative correlation between the dosage administered and the EPS incidence rate (Ereshefsky, 1999). EPS involves the midbrain dopaminergic structure the SN, which has been shown to express Slack channels abundantly. Because these channels are located at dopaminergic neurons, Slack activation would be expected to decrease SN neuron firing and dopamine release. This effect would potentiate the postsynaptic blockade of dopaminergic receptors and would exacerbate EPS. Additional experiments directly addressing the concentration-dependent effects of loxapine and amoxapine metabolites on recombinant Slack channels and native KNa channels would be necessary to test this hypothesis. A study using cultured neurons derived from patients with schizophrenia showed unique and unexplained positive effects of loxapine (10 μM) on neuronal connectivity, compared with structurally related antipsychotic drugs (Brennand et al., 2011). It would be of interest to perform the same experiments in the presence of a Slack channel antagonist or in Slack-knockdown studies, to evaluate the possible involvement of Slack channel activation in this process.
The experiments performed with DRG neurons confirmed that loxapine could activate native KNa channels. Our demonstration that loxapine could affect the excitability of DRG neurons might be important in terms of pain therapeutics. Increasing KNa channel activity should decrease DRG excitability and decrease pain signaling in nociceptive fibers. Although Slack channels are trafficked from the DRG membrane during inflammatory pain signaling (Nuwer et al., 2010), it is not known whether the same process occurs in other painful pathological situations, such as neuropathic pain. Activation of residual KNa channels remaining at the membrane may still offer a way to decrease DRG hyperexcitability; therefore, our studies provide the basis for assessing the putative analgesic properties of Slack channel openers.
This article describes for the first time the Slack channel-activating properties of three marketed molecules, niclosamide, riluzole, and loxapine, which were found by screening the LOPAC with AAS. The efficacy of loxapine has been confirmed for both human and rat recombinant Slack channels and rat KNa channels in DRG neurons. In keeping with the proposed role of Slack channel regulation of DRG neuron excitability, loxapine was found to hyperpolarize the resting potential and to increase the threshold for action potential firing, which suggests that novel analgesics could be developed as Slack channel-targeted drugs. Although the relevance of the effect of loxapine on Slack channels at the clinically achieved dosages is undetermined, the drug can be a valuable tool in characterizing the physiological role of the channel.
Participated in research design: Biton, Sethuramanujam, Bhattacharjee, and Curet.
Conducted experiments: Biton, Sethuramanujam, Picchione, Bhattacharjee, Khessibi, Chesney, and Lanneau.
Wrote or contributed to the writing of the manuscript: Biton, Bhattacharjee, Khessibi, Chesney, Lanneau, Curet, and Avenet.
We thank Noelle Boussac and Laurent Andrieu from the Biostatistic Research and Chemistry Department of Sanofi for their valuable comments and support in the statistical analysis of this work.
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
- regulator of K+ conductance
- atomic absorption spectrometry
- library of pharmacologically active compounds
- extrapyramidal syndrome
- dorsal root ganglion
- substantia nigra
- Chinese hamster ovary
- dimethyl sulfoxide
- human embryonic kidney
- Hanks' balanced salt solution
- α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid
- sodium-activated potassium.
- Received June 9, 2011.
- Accepted December 12, 2011.
- Copyright © 2012 by The American Society for Pharmacology and Experimental Therapeutics