Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Other than neurotransmitter receptors and transporters, ion channels constitute an attractive target to develop new drugs that will be active on the central nervous system. Currently, the only ion channel that is well established as a central nervous system target is the voltage-gated Na⁺ channel, which is blocked by antiepileptic drugs such as phenytoin, carbamazepine, and lamotrigine (McNamara, 1996).

Ca²⁺-activated K⁺ channels play a fundamental role in the control of the firing frequency and pattern of neurons (for reviews, see Sah, 1996; Vergara et al., 1998). Within this class of K⁺ channels, small-conductance voltage-insensitive Ca²⁺-activated K⁺ channels (SK channels) underlie the prolonged postspike afterhyperpolarization (AHP) that is observed in many central neurons. Three closely related SK-channel subtypes, SK1, SK2, and SK3, have been cloned and are characterized by different sensitivities to apamin (Köhler et al., 1996; Strøbæk et al., 2000). These subtypes are differentially expressed in the brain (Stocker and Pedarzani, 2000). Moreover, the currents underlying AHP have been classified into two groups on the basis of their kinetic and pharmacological properties. I_AHP, the current underlying the medium AHP, is present in most excitable cells; it is sensitive to the bee venom toxin apamin, and its activation has been reported to control firing frequency in tonically spiking neurons. sI_AHP, the current underlying the slow AHP, is present in a few cell types only; it is apamin-insensitive but can be modulated by various neurotransmitters, and its activation is responsible for late-spike frequency adaptation (Sah, 1996; Vergara et al., 1998).

Evidence suggests that SK-channel modulation may be interesting in a range of central nervous system disorders, including Alzheimer's disease and schizophrenia. Indeed, apamin has been shown to improve some aspects of memory processing in animal models (Ikonen and Riekkinen, 1999; Fournier et al., 2001), and longer CAG repeats within the gene of the SK3 channel have been associated with schizophrenia in various ethnic groups (Chandy et al., 1998; Dror et al., 1999), although the latter observation is controversial (Antonarakis et al., 1999).

To validate SK channels (or one of their subtypes) as an interesting drug target, it is critical to develop adequate tools for studying their function. Apamin is not ideal because it induces a long-lasting block of SK channels; moreover, its peptidic nature precludes iontophoretic application during in vivo experiments. Several currently available nonpeptidic blockers of SK channels (e.g., (+)-tubocurarine, bicuculline salts, and gallamine) lack specificity, acting on other targets, including GABA_A, nicotinic, or muscarinic receptors (Lee and el-Fakahany, 1991; Dunn et al., 1996; Seutin and Johnson, 1999; Strøbæk et al., 2000). Recently, novel nonpeptidic blockers of SK channels have been synthetised and evaluated. Among these bis-quinolinium cyclophanes, UCL1684 has been found to be the most potent compound, having an IC₅₀ value of 3 nM on AHP of rat superior cervical ganglion neurons (Campos Rosa et al., 2000). However, to our knowledge, a comprehensive screening of various receptors and channels is not available for these compounds, making it difficult to assess their degree of selectivity.

Given the structural analogy between bicuculline and laudanosine, a metabolite of the neuromuscular relaxant atracurium, we decided to evaluate the ability of laudanosine and its methyl, ethyl, butyl, and benzyl derivatives to selectively block SK channels. Both electrophysiological and binding experiments were performed. The electrophysiological study was done on rat midbrain dopaminergic neurons and hippocampus CA1 pyramidal cells. Midbrain dopaminergic neurons display a prominent apamin-sensitive AHP of medium duration, which appears to be mediated by SK3 channels (Shepard and Bunney, 1991; Wolfart et al., 2001). Hippocampus CA1 pyramidal cells present both a medium, apamin-sensitive and a slow, apamin-insensitive AHP (Stocker et al., 1999). SK1 and SK2 subunits are highly expressed in this region (Stocker and Pedarzani, 2000). Binding experiments on SK channels, voltage-gated potassium channels, and muscarinic and nicotinic receptors were performed on rat cerebral cortex membranes.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Electrophysiological Experiments. The methods used were similar to those described previously (Seutin et al., 1997). Male Wistar rats (150-200 g) were used. They were housed and handled in accordance with guidelines of the National Institute of Health (National Institutes of Health Publication 85-23, 1985). Animals were anesthetized with chloral hydrate (400 mg/kg i.p.) and decapitated. The brain was excised quickly and placed in cold (~4°C) artificial cerebrospinal fluid of the following composition 126 mM NaCl, 2.5 mM KCl, 1.2 mM NaH₂PO₄, 1.2 mM MgCl₂, 2.4 mM CaCl₂, 11 mM glucose, and 18 mM NaHCO₃, saturated with 95% O₂ and 5% CO₂ (pH 7.4). A block of tissue containing the midbrain or the hippocampus was cut into horizontal or transverse slices, respectively (thickness 350 µm), in a Vibratome (Lancer, St. Louis, MO). The slice containing the region of interest was placed on a nylon mesh in a recording chamber (volume 500 µl). The tissue was held in position with two electron microscopy grids weighed down by short pieces of platinum wire. The slice was completely immersed in a continuously flowing (~2 ml/min), heated solution (35°C) of the same composition as indicated above. Most recordings of dopaminergic neurons were made from neurons located in the substantia nigra pars compacta. Identification of dopaminergic and CA1 pyramidal cells was performed as described previously (Scuvée-Moreau et al., 1997; Seutin et al., 1997).

Binding Studies. Radioligand binding studies were performed to assess the affinity of laudanosine and its quaternary derivatives for SK channels and the affinity of methyl-laudanosine for voltage-gated potassium channels and muscarinic and nicotinic receptors. Assays were performed by (Cerep/Paris, France) on rat cerebral cortex membranes using general procedures described by Hugues et al. (1982) for SK channels, by Sorensen and Blaustein (1989) for voltage-gated potassium channels, by Richards (1990) for nonselective muscarinic receptors, by Pabreza et al. (1991) for neuronal -bungarotoxin-insensitive nicotinic receptors, and by Sharples et al. (2000) for neuronal -bungarotoxin-sensitive nicotinic receptors.

Compounds. Methyl-laudanosine iodide, ethyl-laudanosine iodide, butyl-laudanosine iodide, and benzyl-laudanosine chloride (Fig. 1) were synthesized in our laboratory according to conventional methods. Briefly, a mixture of laudanosine with an excess of alkyl or aralkyl halide in acetonitrile was refluxed overnight. Excess of reagent and solvent was removed under reduced pressure. For each compound, the crude quaternary salt was isolated in an acetone/diethyl ether mixture. Finally, the products were recrystallized from ethanol, methyl ethyl ketone, diethyl ether/acetone, or dichloromethane/diethyl ether for methyl, ethyl, butyl, and benzyl derivatives, respectively. Purity was checked with classical methods, such as elemental analysis, mass spectrometry, and determination of the melting point.

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Chemical structure of bicuculline, laudanosine, and their quaternary salts, bicuculline methyl iodide, and laudanosine methyl iodide (R = CH3), ethyl iodide (R = C2H5), butyl iodide (R = C4H9), and benzyl chloride (R = CH2C6H5).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Midbrain Dopaminergic Neurons. All recorded neurons (n = 32) displayed characteristic features of identified dopamine neurons (Grace and Onn, 1989), including broad-action potentials (1 ms at 50% maximum height) with a threshold close to 40mV, a prominent I_h current, and a slow AHP, which reached its peak ± 50 ms after the action potential; their mean input resistance was 183 ± 11 M.

View larger version (9K): [in this window] [in a new window]   Fig. 2.   Effect of increasing concentrations of methyl-laudanosine (CH3-L) on the AHP of dopaminergic neurons. A maximal block similar to the one obtained with 300 nM apamin is obtained with 100 µM CH3-L (A). The effect of CH3-L is fully reversible after 10 min in wash (B). Each trace is the mean of four sweeps. Action potentials are truncated.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   A, concentration-response curves for the blockade of the slow AHP in midbrain dopaminergic neurons by laudanosine derivatives. Each data point is presented as the mean ± S.E.M. of three experiments except for methyl- and ethyl-laudanosine (n = 7 and n = 6, respectively). Values for concentrations higher than 300 µM were extrapolated. B, displacement of 125I-apamin binding at rat neocortical SK channels by laudanosine derivatives. Each data point is presented as the mean ± S.E.M. of three experiments. Values for concentrations higher than 1 mM were extrapolated. , Methyl-laudanosine; , ethyl-laudanosine; , benzyl-laudanosine; , butyl-laudanosine; , laudanosine.

View larger version (10K): [in this window] [in a new window]   Fig. 4.   Effect of 100 µM methyl-laudanosine (CH3-L) on cell parameters and GABAA receptors in dopaminergic neurons recorded intracellularly in rat brain slices. A, CH3-L has no effect on spike amplitude or duration. B, CH3-L has no effect on membrane potential, on input resistance measured by the amplitude of the voltage deflection elicited by a hyperpolarizing pulse of 30 pA, or on the Ih current activated by a hyperpolarizing pulse of 120 pA. The experiment was performed in the presence of 0.5 µM tetrodotoxin (TTX); chart speed was reduced during 7 min at the beginning of drug application. C, CH3-L does not modify the decrease in input resistance induced by muscimol, whereas this decrease is fully antagonized by the GABAA antagonist SR95531. Traces are means of voltage deflections induced by a current injection of 60 pA from a potential of 60 mV in control and different treatment conditions. ZD7288 (30 µM) and tetrodotoxin (0.5 µM) were present throughout the experiment.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Effect of methyl-laudanosine and apamin on the excitability of midbrain dopaminergic neurons. A, action potentials (truncated in the figure) were elicited by 200 pA depolarizing current injections before, during, and after application of 100 µM methyl-laudanosine; the membrane potential was set at 60 mV by a continuous injection of 100 pA. B and C, the number of action potentials elicited by 1- to 1.5-s current injections are plotted against the level of intensity of the injected current. Methyl-laudanosine (100 µM) (B) and apamin (300 nM) (C) induced a significant increase in the number of action potentials evoked (, p < 0.05; , p < 0.005). The effect of methyl-laudanosine was reversible (see A). Values are means of eight (methyl-laudanosine) and six (apamin) experiments. Absolute values of current injection were different among experiments; the intensities were chosen to produce increasing numbers of action potentials from 0 to 4 per injection in control conditions.

Hippocampus CA1 Pyramidal Cells. In view of the different regional distribution of SK subunits, it was interesting to investigate the influence of methyl-laudanosine on the excitability of CA1 pyramidal cells, as these cells have a high level of expression for SK1 and SK2 subunits, contrary to midbrain dopaminergic neurons, which express mostly the SK3 subunit (Stocker and Pedarzani, 2000). Experiments were performed on eight cells; their mean input resistance was 55 ± 7.5 M. Under control conditions, long (~1 s) depolarizing pulses elicited trains of action potentials characterized by early- and late-spike frequency adaptation (Fig. 6A, left panel). Methyl-laudanosine (100 µM) failed to significantly modify spiking frequency and pattern (Fig. 6, A and B), whereas apamin (300 nM) induced a significant increase of the number of action potentials evoked without modifying the late-spike frequency adaptation (not shown). The effect of apamin was independent of the amplitude of the injected current (Fig. 6C).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Effect of methyl-laudanosine and apamin on the excitability of hippocampus CA1 pyramidal cells. A, trains of action potentials (truncated in the figure) characterized by early- and late-frequency adaptation were elicited by 150 pA depolarizing current injections; resting membrane potential was 65 mV. Methyl-laudanosine (100 µM) did not modify cell excitability. B and C, the number of action potentials elicited by ~1-s current injections are plotted against the level of intensity of the injected current. Methyl-laudanosine (100 µM) (B) failed to modify the number of action potentials evoked, whereas apamin (300 nM) (C) induced a significant increase in this number (, p < 0.05). Values are means of eight experiments. Absolute values of current injection were different among experiments; the intensities were chosen to produce increasing numbers of action potentials (from 0 to 12) in control conditions.

Binding Studies. The IC₅₀ and K_i values determined for laudanosine, its quaternary derivatives, and the reference compound apamin at neocortical SK channels are presented in Table 1 and Fig. 3B. The order of affinity was methyl-laudanosine > laudanosine  ethyl-laudanosine  benzyl-laudanosine > butyl-laudanosine. The IC₅₀ and K_i values determined for methyl-laudanosine and the reference compounds at voltage-gated potassium channels and muscarinic and nicotinic receptors are shown in Table 2. Depending on the parameter being considered, the affinity of methyl-laudanosine for SK channels was 8 to 30 times superior to its affinity for muscarinic receptors. The affinity for nicotinic receptors was weak or negligible depending upon the subtype examined (-bungarotoxin-insensitive or -sensitive). No affinity for voltage-gated potassium channels could be detected.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This study was initiated to identify a specific and reversible blocker of SK channels. Drawbacks of blockers such as apamin, bicuculline, (+)-tubocurarine, and gallamine have been stressed earlier (see the Introduction).

Our results show that quaternary laudanosine derivatives differentially block the slow AHP in rat dopaminergic neurons, with the N-methyl derivative being the most potent blocker. This property, which was previously described for quaternary salts of bicuculline, namely, bicuculline methyl iodide and bicuculline methochloride (Seutin et al., 1997), was suspected given the similarity of structure between these compounds. Methyl-laudanosine is slightly more potent as a blocker of the AHP than bicuculline salts, whose IC₅₀ was found to be 26 µM in the same conditions (Seutin et al., 1997). The difference of potency (~10 times) between the methyl derivative of laudanosine and the base is similar to that reported previously between the quaternary salts of bicuculline and the base (Seutin et al., 1997). Increasing the length of the N-substituent does not increase the activity on SK channels.

Experiments on cell excitability show that apamin and methyl-laudanosine significantly increase the number of spikes evoked in midbrain dopaminergic neurons by positive current injection. These results are consistent with other studies indicating a role for SK channels in the control of the firing frequency of these cells (Shepard and Bunney, 1991; Wolfart et al., 2001). Interestingly, an increase in the firing frequency of hippocampus CA1 pyramidal cells was observed with apamin but not with methyl-laudanosine. Our results with apamin are in agreement with previous studies indicating a contribution of apamin-sensitive SK channels to the firing properties of hippocampal pyramidal neurons (Stocker et al., 1999). The fact that methyl-laudanosine does not modify cell excitability at a concentration inducing a maximal block of the AHP in dopaminergic neurons suggests the possibility that this compound discriminates between SK subtypes. In fact, recent studies show a high level of expression of SK3 subunits in midbrain dopaminergic neurons (Tacconi et al., 2001; Wolfart et al., 2001) and a high level of expression of SK1 and SK2 subunits in CA1 to CA3 pyramidal neurons (Stocker et al., 1999). Electrophysiological results thus point to a preferential effect of methyl-laudanosine on the SK3 subtype of SK channels. However, further studies in slices and in cell lines expressing the various subunits are needed to address this point thoroughly.

Binding data obtained from rat cerebral cortex membranes are in good agreement with the electrophysiological study on dopaminergic cells and confirm the good affinity of methyl-laudanosine for SK channels. In fact, IC₅₀ values obtained in the two kinds of studies are quite similar (~15 µM in electrophysiological studies and ~4 µM in binding experiments). Furthermore, the difference of potency (~10 times) between laudanosine and its methyl derivative is also similar in the two studies. With the exception of ethyl-laudanosine, the order of affinity for SK channels among laudanosine derivatives (methyl-laudanosine > laudanosine  ethyl-laudanosine  benzyl-laudanosine > butyl-laudanosine) is comparable with their order of potency as AHP blockers (methyl-laudanosine > ethyl-laudanosine > laudanosine > benzyl-laudanosine > butyl-laudanosine). As pointed out earlier, this may be due to the differences in SK subtypes that exist between brain regions.

Additional data from our study show that methyl-laudanosine has no affinity for GABA_A and nicotinic receptors. This is an interesting finding because pharmacological studies suggest that the tridimensional structure of the binding site in SK, GABA_A, and nicotinic channels is similar (Seutin and Johnson, 1999). Furthermore, methyl-laudanosine is less potent at muscarinic receptors than at SK channels. Indeed, more recent iontophoretic experiments performed in vivo in our laboratory suggest that local effects of methyl-laudanosine are not due to muscarinic effects (Seutin V., Massotte L., Liégeois J.-F., and Scuvée-Moreau J., unpublished results). Finally, both our electrophysiological and binding studies suggest that methyl-laudanosine does not interact with fast Na⁺ channels, voltage-dependent K⁺ channels, or I_h channels. It should be noted, however, that we cannot exclude an action of the compound on another receptor or channel.

Methyl-laudanosine has a relatively low potency when compared with apamin or UCL1684. This is probably due to a high off-rate of the compound from the channel that makes it a rapidly reversible blocker. This property may be interesting for pharmacological experiments in which a relatively quick reversal of the effect is needed. Again, our recent in vivo iontophoretic experiments show that this is the case.

Our study demonstrates the pharmacological differences between bicuculline quaternary salts and methyl-laudanosine. Additional modifications of this alkaloid will be performed to further study the structural elements required for a putative selective action on SK-channel subtypes.

In conclusion, the lack of influence of methyl-laudanosine on various neuronal parameters combined with its lack of antagonism at GABA_A receptors and its lack of affinity for nicotinic receptors suggest that this agent may become an interesting tool to examine the functional role of SK channels in various experimental conditions.