Inhibition of voltage-gated calcium channels by fluoxetine in rat hippocampal pyramidal cells

Neuropharmacology 39 (2000) 1029–1036 www.elsevier.com/locate/neuropharm

Inhibition of voltage-gated calcium channels by fluoxetine in rat hippocampal pyramidal cells Ferenc Dea´k a, Ba´lint Laszto´czi a, Pa´l Pacher b, Ga´bor L. Petheo¨ a, Vale´ria Kecskeme´tib, Andra´s Spa¨t a,* a

Department of Physiology, Laboratory of Cellular and Molecular Physiology, Semmelweis University of Medicine, P.O. Box 259, H-1444 Budapest, Hungary b Department of Pharmacology, Semmelweis University of Medicine, P.O. Box 370, H-1445 Budapest, Hungary Accepted 12 August 1999

Abstract Fluoxetine, an antidepressant which is used world-wide, is a prominent member of the class of selective serotonin re-uptake inhibitors. Recently, inhibition of voltage-gated Na+ and K + channels by fluoxetine has also been reported. We examined the effect of fluoxetine on voltage-gated calcium channels using the patch-clamp technique in the whole-cell configuration. In hippocampal pyramidal cells, fluoxetine inhibited the low-voltage-activated (T-type) calcium current with an IC50 of 6.8 µM. Fluoxetine decreased the high-voltage-activated (HVA) calcium current with an IC50 between 1 and 2 µM. Nifedipine and ωconotoxin GVIA inhibited the HVA current by 24% and 43%, respectively. Fluoxetine (3 µM), applied in addition to nifedipine or ω-conotoxin, further reduced the current. When fluoxetine (3 µM) was applied first neither nifedipine nor ω-conotoxin attenuated the remaining component of the HVA current. This observation indicates that fluoxetine inhibits both L- and N-type currents. In addition, fluoxetine inhibited the HVA calcium current in carotid body type I chemoreceptor cells and pyramidal neurons prepared from prefrontal cortex. In hippocampal pyramidal cells high K +-induced seizure-like activity was inhibited by 1 µM fluoxetine; the mean burst duration was shortened by an average of 44%. These results provide evidence for inhibition of T-, N- and L-type voltage-gated calcium channels by fluoxetine at therapeutically relevant concentrations.  2000 Elsevier Science Ltd. All rights reserved. Keywords: Depression; Epilepsy; Fluoxetine; Voltage-gated calcium channel; Hippocampus; Pyramidal neuron

1. Introduction Fluoxetine is a widely used antidepressant compound, its action is primarily attributed to inhibition of the reuptake of serotonin in the central nervous system (Wong et al., 1995). Recent studies have indicated, however, that fluoxetine has several additional effects, including the blockade of ion channels. High concentrations (10– 100 µM) of fluoxetine have been reported to block voltage-gated Na+ (Pancrazio et al., 1998) and K + channels (Tytgat et al., 1997) in neurons and in epithelial cells, cultured from human cornea and lens (Rae et al., 1995). In smooth muscle cells, low concentrations of fluoxetine

* Corresponding author. Tel.: +36-1-266-9180; fax: +36-1-2666504. E-mail address: [email protected] (A. Spa¨t).

(0.1–10 µM) decreased the delayed rectifier K + current, while at higher concentrations (above 100 µM) the drug increased the calcium-activated K + current (Farrugia, 1996). Fluoxetine was found to inhibit various Cl⫺ channels (Maertens et al., 1999). Fluoxetine decreased K +-induced calcium transients in synaptosomes (Stauderman et al., 1992) suggesting that the drug inhibits voltage-gated calcium channels. This was affirmed in a rat tumor cell-line (Hahn et al., 1999), however, the channel type inhibited by the drug was not identified in these studies. The aim of the present study was to provide direct evidence for the inhibitory effect of fluoxetine on the voltage-gated calcium channels in neurons, using the patch-clamp method. As some antiepileptic drugs are known to inhibit voltage-gated calcium channels, a supposed anticonvulsant action of fluoxetine (Leander, 1992) was tested in the high K + model of epilepsy.

0028-3908/00/$ - see front matter  2000 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 8 - 3 9 0 8 ( 9 9 ) 0 0 2 0 6 - 3

1030

F. Dea´k et al. / Neuropharmacology 39 (2000) 1029–1036

2. Methods 2.1. Cell preparation and culture Hippocampal and prefrontal cells were prepared and cultured as described (Dea´k et al., 1998), with minor modification. Briefly, the hippocampus and frontal pole of the cortex were obtained from decapitated rat embryos at embryonic days 18 to 20. The embryos were removed from pregnant Wistar rats, which were decapitated with a guillotine after striking the cranium. Cells were isolated by mechanical dissociation and plated onto the poly-d-lysine-coated surface of a Petri dish and cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum, 200 IU/ml penicillin and 200 µg/ml streptomycin. The potassium concentration [K +] of the culture medium was 5.3 mM. Cultures were maintained at 37°C under 5% CO2 for one to three weeks. Type I chemoreceptor cells were enzymatically isolated from the carotid bodies of 14–21-day-old rat pups. Two to four animals were anesthetized with sodium pentobarbital (5 mg intraperitoneal). The head together with the neck was cut off and placed in ice-cold saline. The carotid artery at the bifurcation was excised under a stereomicroscope and placed in ice-cold phosphatebuffered saline (PBS). The carotid bodies were removed and incubated in low-Ca2+ (0.2 mM) PBS containing type 1 collagenase (2 mg/ml) and trypsin (2 mg/ml) at 37°C for 17 min. After teasing apart, the carotid bodies were digested for a further 5 min. Following mechanical dissociation cells were centrifuged for 10 min at 200 g. The pellet was re-suspended in culture medium (containing Dulbecco’s modified Eagle’s medium and Ham’s F-12 (1:1) and supplemented with 10% heat inactivated fetal calf serum, 100 IU/ml penicillin, 100 µg/ml streptomycin and 84 IU/l insulin). Twenty microliter aliquots were plated onto 35 mm plastic Petri dishes coated with poly-d-lysine and 2 h were allowed for adhesion in an incubator under 5% CO2 at 37°C. Two milliliters of culture medium was added to Petri dishes and thereafter cells were cultured for 24–50 h. Phase-bright chemoreceptor cells with polygonal shape and short processes had good seal-forming ability. The experiments were carried out in accordance with the guidelines of the European Communities Council Directive (86/609/EEC) and were approved by the Animal Care and Ethics Committee of the Semmelweis University (approval No. 124/94 and 17-3/98). 2.2. Electrophysiological recordings Membrane potential and whole-cell membrane currents were recorded from all cell types using the patchclamp technique from the somatic region of cells chosen on a morphological basis under visual control with Nikon Axiophot inverted microscope. For electrophysi-

ological measurements the culture medium was changed to bath solution 1 (see below), and the cells were maintained at 30°C. Test solutions were applied by a gravity driven system within 100 µm of the cell. To protect from light, tubes containing light-sensitive drugs like fluoxetine or nifedipine, were wrapped in opaque material. Patch pipettes were pulled using borosilicate glass (GC120F-10, Clarke Electromedical, Pangburne, Reading, UK) with a P-87 micropipette puller (Sutter Instruments, Novato, CA, USA). Pipette resistance was 2–9 M⍀ when filled with P1 (see below). Holding potential was set to ⫺100 mV unless otherwise stated. Records were not corrected for either series resistance or for leak unless otherwise stated. Electrical signals were amplified with an Axopatch 1D amplifier (Axon Instruments, Foster City, CA, USA) or with a RK-400 patch-clamp amplifier (Biologic Science Instruments, Claix, France), low-pass filtered at 1–2 kHz with an eight-pole Bessel filter and digitized on-line using a Digidata 1200 interface (Axon Instruments, Foster City, CA, USA) and pClamp6 software (Axon Instruments, Foster City, CA, USA) at 5–20 kHz. The data were stored on IBM-compatible computer hard disk. 2.3. Data analysis Data analysis was performed using pClamp6 software. For the high K + model of epilepsy, depolarization shifts longer than 120 ms were statistically evaluated in burst analysis. Results are presented as means±S.E.M. Statistical significance was estimated by Student’s paired-sample t-test, or one-way analysis of variance and Newman– Keuls test, as appropriate. A P-value less than 0.05 was considered statistically significant. 2.4. Solutions (in mM): Bath 1 (B1): 140 NaCl, 2 KCl, 2 CaCl2, 0.5 MgCl2, 11 glucose, 10 hydroxyethylpiperazine-N⬘-2-ethanesulfonic acid (HEPES), pH 7.4. When tetraethylammonium (TEA, 25 mM) was added to B1, isotonicity was maintained by equimolar reduction of Na+. Bath 2 (B2): 25 TEA-Cl, 115 N-methyl-d-glucamine (NMDG)-Cl, 2 KCl, 2 CaCl2, 0.5 MgCl2, 0.1 BaCl2, 11 glucose, 10 HEPES, pH 7.4. Pipette solution 1 (P1): 110 K-gluconate, 25 KCl, 0.05 CaCl2, 2 MgCl2, 1 ethyleneglycol-bisN,N,N⬘,N⬘-tetraacetic acid (EGTA), 2 Na2ATP, 10 HEPES, pH 7.3. Pipette solution 2 (P2): 135 CsCl, 0.05 CaCl2, 2 MgCl2, 1 EGTA, 2 Na2ATP, 10 HEPES, pH 7.3. In pipette solutions the calculated free [Ca2+] was below 10⫺8 M. Solutions used in experiments with chemoreceptor cells can be read in the legend to Fig. 6.

F. Dea´k et al. / Neuropharmacology 39 (2000) 1029–1036

1031

2.5. Materials The same results were obtained when fluoxetine ((±)N-methyl-3-phenyl-3-((α,α,α-trifluoro-p-tolyl)-oxy)propylamine hydrochloride) from RBI (Natick, MA, USA) or Fujimoto Co. (Osaka, Japan) was used. Culture medium and insulin were purchased from Gibco, foetal calf serum from Protein GMK (Go¨do¨llo¨, Hungary), ωconotoxin GVIA and tetrodotoxin from Alomone Labs (Jerusalem, Israel). All other chemicals were from Sigma.

3. Results 3.1. Inhibition of voltage-gated currents by fluoxetine in hippocampal pyramidal cells To examine the effect of fluoxetine on the voltagegated calcium channels we used voltage-clamped hippocampal pyramidal neurons. After elimination of the sodium current using tetrodotoxin and/or sodium-free bath solution B2, ramp depolarization revealed two inward currents (Fig. 1). As we reported previously (Dea´k et al., 1998), the low-voltage-activated (LVA) current corresponds to the T-type calcium current. The high-voltage-activated (HVA) current may be attributed to the activation of N-type and, to a smaller extent, also of L-type Ca2+ channels. In the present study, a T-type current was seen in 58 of 117 cells evaluated, while HVA current was present in all cells. Fluoxetine at 3 µM concentration reduced the voltagegated calcium currents (VGCC), as shown during ramp depolarization from ⫺100 mV to +60 mV (Fig. 1). About 1 min was necessary for full inhibition and it was partially or not at all reversible during the subsequent 3min wash-out period. To analyze the effect of fluoxetine on the LVA calcium current, step depolarization from ⫺100 mV to ⫺40 mV was applied at every 15 s (Fig. 2). This current was sensitive to 100 µM NiCl2, that decreased the LVA current by 84±16% (n=3). The activation at this rather negative potential, fast inactivation (τ=25±2 ms at ⫺40mV, n=6) and Ni2+ sensitivity correspond to the T-type calcium current. Fluoxetine inhibited the T-type current in a concentration-dependent manner with an IC50=6.8 µM (Fig. 2B). Fluoxetine (3 µM) reduced the decay time constant to τ=18±2 ms (P⬍0.05, n=6), which indicates an increased rate of T-type channel inactivation in the presence of the drug. To quantify the inhibition of the HVA Ca2+ channel by fluoxetine, T-type channels were inactivated with a holding potential of ⫺60 mV and a step depolarization to ⫺10 mV was applied every 15 s (Fig. 3A). Consecutive application of 0.3 µM and higher concentrations of fluoxetine significantly reduced the HVA calcium cur-

Fig. 1. The effect of nifedipine and fluoxetine on the high-voltagegated calcium currents in hippocampal pyramidal cells. In voltageclamp experiments ramp depolarization from ⫺100 mV to +60 mV (875 mV/s) was applied. Note the further reduction of the calcium currents by fluoxetine (3 µM) compared to the effect of nifedipine (2 µM). Incubation medium B2 and pipette solution P2 (see Solutions in Methods) were used. These traces are representative of four experiments.

rent with an IC50=1.1 µM (n=9, Fig. 3B). Run-down was determined for these experiments, and the average reduction rate of the peak current was 2.7% min⫺1 (n=9). Depending on the duration of measurement, error due to run-down was between 5 and 20% of the current, for instance, after 5 min the current decreased to 84±5% of its original amplitude. If we correct for run-down, the IC50 of the HVA current inhibition by fluoxetine is still below 2 µM. Inhibition of the HVA calcium current by fluoxetine (3 µM) was voltage independent in the voltage range of ⫺20 to +30 mV. The reduction was 56±10% at ⫺20 mV and 56±8% at +30 mV (n=8, P⬎⬎0.05). Nifedipine, a dihydropyridine antagonist of L-type calcium channels, reduced the peak amplitude of HVA current by 24±2% (n=22, P⬍0.001, Fig. 4A). Nifedipine had no significant effect in the presence of 3 µM fluoxetine (inhibition by 58±15% with fluoxetine vs 70±11% with fluoxetine and nifedipine, n=4, P⬎0.1), which

1032

F. Dea´k et al. / Neuropharmacology 39 (2000) 1029–1036

Fig. 2. Reduction of the T-type calcium current with increasing concentrations of fluoxetine, applied consecutively to the same cell. Panel A shows the T-type calcium current induced with step depolarization from ⫺100 mV to ⫺40 mV. Records were corrected for leak and capacitive currents. For clarity the first 5 ms after step to ⫺40 mV were omitted. B2 and P2 were used. Traces are representative of five cells, except for the highest concentration (30 µM), which was tested in two cells only. Panel B depicts the mean peak current ±S.E.M. of T-type calcium currents at ⫺40 mV in the function of the concentration of fluoxetine. The curve was fitted to the equation: Inhibition (%)=100/[1+(Ki/C)q], in which C is the drug concentration, q the Hill coefficient and Ki the drug concentration needed for half-maximal block. Ki was 6.75 µM and q=1.32 (r=0.82).

Fig. 3. Inhibition of the HVA calcium current with increasing concentrations of fluoxetine. In panel A the calcium current was activated by step depolarization from ⫺60 mV to ⫺10 mV. For clarity the first 5 ms after step to ⫺10 mV have been omitted. B2 and P2 were used. The figure shows representative traces after correction for leak and capacitance currents for five to 12 cells. Panel B shows the mean peak current ±S.E.M. (n=4–12 cells) of HVA calcium current at ⫺10 mV in the presence of increasing concentrations of fluoxetine. The curve was fitted to the same equation as in Fig. 2. In this case Ki was 1.07 µM and q=0.73 (r=0.79).

means a complete or nearly complete inhibition of Ltype Ca2+ channels by fluoxetine. This observation was further analyzed by addition of fluoxetine in the presence of ω-conotoxin GVIA, a selective antagonist of N-type calcium channels (Fig. 4B). Adding 1.5 µM of ω-con-

otoxin GVIA reduced the HVA current by 43±5% (n=11), which indicates the significant contribution of N-type channels to the HVA current. Fluoxetine (3 µM) further decreased this remaining current by 25±7% (P⬍0.01, n=6), which may be attributed to the inhibition of the L-type channel by fluoxetine.

F. Dea´k et al. / Neuropharmacology 39 (2000) 1029–1036

1033

Fig. 4. Inhibition of L- and N-type calcium channels by fluoxetine. Step depolarization from ⫺60 mV to ⫺10 mV was applied every 15 s. (A) Addition of nifedipine (2 µM) was followed by combined application of this drug and fluoxetine (3 µM). The traces are representative of six cells. (B) Application of ω-conotoxin GVIA (1.5 µM) was started first and continued together with that of fluoxetine (3 µM). The traces are representative of six cells. In both cases traces were corrected by subtracting a trace recorded in the presence of 100 µM CdCl2. Solutions B2, completed with 2 mM 4-aminopyridine and 0.5 µM tetrodotoxin, and P2 were used.

Addition of 3 µM fluoxetine resulted in 55±9% (P⬍0.001, n=9) inhibition of the HVA calcium current. Considering that the fluoxetine-sensitive fraction of the current is much bigger than the nifedipine-sensitive fraction, the N-type calcium current is certainly inhibited by fluoxetine. To prove this, fluoxetine (3 µM) was applied in the presence of nifedipine (Fig. 1 and Fig. 4B), which resulted in a further reduction of the peak amplitude of the HVA current to 40±7% of the control (P⬍0.001, n=6). Furthermore, adding ω-conotoxin GVIA after fluoxetine (3 µM) had no effect (43±7% of the control, P⬎⬎0.05, n=9). As fluoxetine eliminates the inhibition of the HVA current by ω-conotoxin, fluoxetine probably also completely inhibits the N-type calcium channel. Using another N-type Ca2+ channel antagonist ω-conotoxin MVIIA at a concentration of 2 µM gave slightly higher inhibition of the HVA current but a similar interaction with fluoxetine. 3.2. Anticonvulsant effect of fluoxetine in the high K+model of epilepsy Using the patch-clamp current-clamp technique, spontaneous synaptic activity was recorded from hippocampal cells forming neuronal networks in culture. In control solution action potentials occurred spontaneously rather rarely. Application of 7 mM K + in the bath switched the spontaneous synaptic activity to burst firing, for which a depolarization shift and action potentials superimposed on it were characteristic (Fig. 5). This effect of high K + was seen even when hyperpolarizing current was manually applied. Therefore K +-induced depolarization,

Fig. 5. Seizure-like activity in cultured hippocampal neural network is attenuated by fluoxetine (1 µM). Solution B1 was completed with 5 mM KCl or NMDG-Cl (control). The depolarization caused by K + was cancelled out with manually clamping the cell near to its resting membrane potential. Solution P1 was used.

changing electrochemical gradient of Ca2+ or inducing inactivation of voltage-gated calcium channels, could be avoided. We tested the effect of fluoxetine at the therapeutically relevant 1 µM concentration on burst firing in 13 cells, from which 11 responded to the drug. Fluoxetine did not influence the mean burst frequency (70±14/min) but it reduced the mean burst duration (848±248 ms) to 56±4% (n=11, P⬍0.001). In three separate experiments the burst firing was induced by combined application of tetraethylammonium (25 mM) and K + (12 mM) in the bath. Mean burst frequency was 15.9±0.3/min, the mean burst duration was 1 117±651 ms. The membrane potential was set with manual clamping also in this case. The burst firing was suppressed with 1 µM fluoxetine, which shortened the bursts to 49±7% (n=3, P⬍0.01) with an unchanged frequency. The anticonvulsive effect of some drugs are thought to involve the inhibition of the fast sodium channel. Therefore, we examined the effect of fluoxetine on the fast, tetrodotoxin-sensitive sodium current under conditions used for high potassium-evoked seizure model. Fluoxetine applied at 1 µM had no effect on the voltagegated sodium current, and at 10 µM it reduced the cur-

1034

F. Dea´k et al. / Neuropharmacology 39 (2000) 1029–1036

rent by 15±2% (P⬍0.01, n=3) (Fig. 6). The HVA calcium current was markedly reduced by 10 µM fluoxetine also in these experiments. 3.3. Inhibition of VGCCs by fluoxetine in prefrontal neurons and in carotid body type I chemoreceptor cells To see whether VGCCs in cells from different regions of the central and peripheral nervous system are also sensitive to fluoxetine, we examined two other cell types. Fluoxetine was an effective inhibitor of VGCCs in prefrontal neurons. Applying at 3 µM, fluoxetine reduced the HVA calcium current, activated by step depolarization from ⫺60 mV to ⫺10 mV, by 43±2% (P⬍0.005, n=3, Fig. 7). Carotid body type I chemoreceptor cells are involved in the regulation of ventilation and are sensitive to hypoxia, hypercapnia and acidosis. Ramp depolarization revealed a HVA current, which was inhibited more than 50% by 2 µM nifedipine (unpublished observation of G. Petheo¨, Z. Molna´r, and A Spa¨t). Fluoxetine (3 µM) inhibited the HVA calcium current by 53±2% (P⬍0.001, n=4, Fig. 8). After wash out of fluoxetine the HVA current returned to 61±5% of its original amplitude. These observations are comparable to the effect of fluoxetine in the hippocampal pyramidal cells.

Fig. 7. The inhibitory effect of fluoxetine on the HVA calcium current in prefrontal pyramidal neurons. Fluoxetine at 3 µM reduced the HVA calcium current activated with step depolarization from ⫺60 mV to ⫺10 mV. Solutions B2 and P2 were used. For clarity the first 5 ms after the step to ⫺10 mV were omitted. Leak and capacitive currents were subtracted from the traces, which are representative of three experiments.

4. Discussion 4.1. Fluoxetine as voltage-gated calcium channel inhibitor

Fig. 6. The effect of a high concentration of fluoxetine on the voltage-gated sodium and calcium currents. Note the strong reduction of HVA calcium current by 10 µM fluoxetine during ramp depolarization. Solutions B1, completed with 25 mM TEA, 100 µM BaCl2 and 10 mM KCl, and P1 were used. Traces are representative of three experiments.

In this study we have shown that fluoxetine, already at micromolar concentration, inhibits L-, N- and T-type calcium channels in hippocampal pyramidal neurons. To our knowledge, this is the first electrophysiological evidence for the inhibition of calcium channels by fluoxetine in hippocampal neurons. Inhibition of VGCCs by fluoxetine has been observed also in prefrontal pyramidal neurons as well as in the carotid body type I chemoreceptor cells. Our data show that the inhibitory effect of fluoxetine on HVA calcium current, at least in hippocampal pyramidal cells, is much greater than inhibition of L-type current by nifedipine. Considering that about the half of HVA calcium current in these cells belongs to the Ntype current and no P-type current has been detected (Dea´k et al., 1998), as well as the fact that in the presence of fluoxetine (3 µM) neither nifedipine nor ω-conotoxins attenuated the remaining component of the HVA current, the conclusion can be drawn that fluoxetine certainly inhibits both the L- and N-type calcium channels. This inhibition is very potent, the value of the IC50 is less than 2 µM. A higher value was presented for the fluoxetine inhibition of calcium uptake in synaptosomes

F. Dea´k et al. / Neuropharmacology 39 (2000) 1029–1036

1035

can be accumulated in tissues, and twenty-fold the plasma concentration of the drug was detected in human brain during chronic fluoxetine treatment (Karson et al., 1993; Komoroski et al., 1994). In view of these data, a significant reduction of the VGCCs by fluoxetine may occur in patients chronically treated with fluoxetine. 4.2. Anticonvulsant effect of fluoxetine in the high K+model of epilepsy

Fig. 8. Calcium currents are reduced by fluoxetine in carotid body type-I chemoreceptor cells. Ramp depolarization from ⫺100 mV to +60 mV (437 mV/s) was used. The extracellular solution contained (mM): N-methyl-d-glucamine-Cl 125, CsCl 1, TEA-Cl 5, MgCl2 5, CaCl2 2, HEPES 3.3, PIPES 3.3, Mes 3.3, pH 7.4. The pipette solution contained (mM): CsCl 135, MgCl2 1, BAPTA 2, ATP 1, HEPES 10, pH 7.2. Traces are representative of four experiments.

(IC50=27 µM, Lavoie et al., 1997) and that of calcium channels in PC12 cells (IC50=13 µM, Hahn et al., 1999). The discrepancy between IC50 values may be attributed to the differences in the parameter examined (45Ca2+ uptake or ion current). Since synaptosomes and PC12 cells may contain P/Q- and R-type calcium channels as well and the action of fluoxetine on these channels has not been characterized, a different contribution of such channels to the total calcium current may also modify the IC50 of fluoxetine. Fluoxetine also attenuated the Ttype calcium current by increasing the inactivation rate of the current. In view of these data and previous reports on the fluoxetine evoked vasodilation (Pacher et al., 1999), inhibition of K +-induced uterine contraction (Velasco et al., 1997) and nitrendipine binding to synaptosomal dihydropyridine receptors (Stauderman et al., 1992), it appears that fluoxetine non-selectively inhibits at least three (L-, N- and T-) types of voltage-gated calcium channels. The lowest fluoxetine concentration with a significant effect on the VGCCs in this study (0.3 µM) falls well within the range of therapeutic plasma concentrations (0.15–1.5 µM) (Orsulak et al., 1988; Kelly et al., 1989; Pato et al., 1991). However, under certain conditions, e.g., decreased elimination in the elderly, the plasma concentration of fluoxetine can reach higher levels (Pato et al., 1991; Borys et al., 1992). In addition, fluoxetine

Several conventional anticonvulsant drugs (e.g., ethosuximide, phenytoin and valproate) were found to inhibit VGCCs (for review see Macdonald and Kelly, 1995). Dihydropyridine antagonists were also claimed to reduce epileptiform activity (as reviewed by Stefani et al., 1997). Our observation, that fluoxetine acts as a calcium channel antagonist and not only as a serotonin transporter inhibitor, gives rise to the possibility that fluoxetine attenuates seizures. Elevation of [K +]o to 7–10.5 mM evokes seizure in hippocampal slices (Traynelis and Dingledine, 1988; Chamberlin and Dingledine, 1989) and this high K +-induced seizure serves as a model for epilepsy (for review see McNamara, 1994). Fluoxetine attenuated bursting in the high K +-induced model for epileptic electrographic seizures. Inhibition of serotonin re-uptake is very unlikely to account for the anticonvulsive effect of fluoxetine under our experimental conditions. The serotonergic neurons are located in the raphe nuclei, and are not included in our preparation. Continuous local perfusion also prevented the accumulation of any serotonin around the cells tested. Inhibition of voltage-gated sodium channels may also reduce epileptiform activity (Macdonald and Kelly, 1995). In our experiments 10, rather than 1 µM fluoxetine was capable of reducing voltage-gated sodium currents. Therefore an effect of fluoxetine on sodium channels may not explain the shortening of bursts in the high-potassium model of epilepsy. Several reports indicate that VGCCs significantly contribute to the increased neuronal activity and concomitant depolarization shift during epileptiform activity (for review see Speckmann et al., 1993). Dihydropyridine derivatives influence the depolarization shift in epileptic foci (Walden et al., 1986; Witte et al., 1987). With all these data in mind, the inhibition of VGCCs by fluoxetine may account for the anticonvulsant action of the drug. Anticonvulsant action of fluoxetine was observed both in animal and human studies. In animal models of epilepsy fluoxetine decreased the probability of seizure and suppressed the electrographic seizure (Prendiville and Gale, 1993; Wada et al., 1995; Dailey et al., 1996). Fluoxetine enhanced the effects of various conventional anticonvulsants in mouse (Leander, 1992). These results are in accordance with clinical reports on lower occurrence of seizures in chronic epileptic patients after co-

1036

F. Dea´k et al. / Neuropharmacology 39 (2000) 1029–1036

administration of fluoxetine with phenobarbital, carbamazepine or valproate (Favale et al., 1995). Our observations on the inhibition of the VGCCs and the attenuation of high K +-induced seizures by fluoxetine may provide new information on the mechanism underlying the action of this drug.

Acknowledgements The expert technical assistance of Ms Krisztina Na´dasy and Ms Aniko´ Rajki is highly appreciated. This work was supported by grants from the Hungarian National Science Foundation (OTKA grant No. T-26173, T19206 and T-14649) and Hungarian Academy of Sciences (No. 97-16 3,2).

References Borys, D.J., Setzer, S.C., Ling, L.J., Reisdorf, J.J., Day, L.C., Krenzelok, E.P., 1992. Acute fluoxetine overdose: a report of 234 cases. American Journal of Emergency Medicine 10, 115–120. Chamberlin, N.L., Dingledine, R., 1989. Control of epileptiform burst rate by CA3 hippocampal cell afterhyperpolarizations in vitro. Brain Research 492, 337–346. Dailey, J.W., Yan, Q.S., Adams Curtis, L.E., Ryu, J.R., Ko, K.H., Mishra, P.K., Jobe, P.C., 1996. Neurochemical correlates of antiepileptic drugs in the genetically epilepsy-prone rat (GEPR). Life Sciences 58, 259–266. Dea´k, F., Nagy, G., Va´rnai, P., Madara´sz, E., Spa¨t, A., 1998. Calcium current activated by potassium ions in voltage-clamped rat hippocampal pyramidal neurones. Journal of Physiology, London 508, 735–745. Farrugia, G., 1996. Modulation of ionic currents in isolated canine and human jejunal circular smooth muscle cells by fluoxetine. Gastroenterology 110, 1438–1445. Favale, E., Rubino, V., Mainardi, P., Lunardi, G., Albano, C., 1995. Anticonvulsant effect of fluoxetine in humans. Neurology 45, 1926–1927. Hahn, S.J., Choi, J.S., Rhie, D.J., Oh, C.S., Jo, Y.H., Kim, M.S., 1999. Inhibition by fluoxetine of voltage-activated ion channels in rat PC12 cells. European Journal of Pharmacology 367, 113–118. Karson, C.N., Newton, J.E., Livingston, R., Jolly, J.B., Cooper, T.B., Sprigg, J., Komoroski, R.A., 1993. Human brain fluoxetine concentrations. Journal of Neuropsychiatry and Clinical Neuroscience 5, 322–329. Kelly, M.W., Perry, P.J., Holstad, S.G., Garvey, M.J., 1989. Serum fluoxetine and norfluoxetine concentrations and antidepressant response. Therapeutics and Drug Monitoring 11, 165–170. Komoroski, R.A., Newton, J.E., Cardwell, D., Sprigg, J., Pearce, J., Karson, C.N., 1994. In vivo 19F spin relaxation and localized spectroscopy of fluoxetine in human brain. Magnetic Resonance in Medicine 31, 204–211. Lavoie, P.A., Beauchamp, G., Elie, R., 1997. Atypical antidepressants inhibit depolarization-induced calcium uptake in rat hippocampus synaptosomes. Canadian Journal of Physiology and Pharmacology 75, 983–987. Leander, J.D., 1992. Fluoxetine, a selective serotonin-uptake inhibitor,

enhances the anticonvulsant effects of phenytoin, carbamazepine, and ameltolide (LY201116). Epilepsia 33, 573–576. Macdonald, R.L., Kelly, K.M., 1995. Antiepileptic drug mechanisms of action. Epilepsia 36 (Suppl. 2), S2–S12. Maertens, C., Wei, L., Voets, T., Droogmans, G., Nilius, B., 1999. Block by fluoxetine of volume-regulated anion channels. British Journal of Pharmacology 126, 508–514. McNamara, J.O., 1994. Cellular and molecular basis of epilepsy. Journal of Neuroscience 14, 3413–3425. Orsulak, P.J., Kenney, J.T., Debus, J.R., Crowley, G., Wittman, P.D., 1988. Determination of the antidepressant fluoxetine and its metabolite norfluoxetine in serum by reversed-phase HPLC with ultraviolet detection. Clinical Chemistry 34, 1875–1878. ´ ., 1999. Serotonine Pacher, P., Ungva´ri, Z., Kecskeme´ti, V., Koller, A reuptake inhibitor, fluoxetine, dilates isolated skeletal muscle arterioles. Possible role of altered Ca2+ sensitivity. British Journal of Pharmacology 127, 740–746. Pancrazio, J.J., Kamatchi, G.L., Roscoe, A.K., Lynch, C. IIIrd, 1998. Inhibition of neuronal Na+ channels by antidepressant drugs. Journal of Pharmacology and Experimental Therapeutics 284, 208–214. Pato, M.T., Murphy, D.L., DeVane, C.L., 1991. Sustained plasma concentrations of fluoxetine and/or norfluoxetine four and eight weeks after fluoxetine discontinuation. Journal of Clinical Psychopharmacology 11, 224–225. Prendiville, S., Gale, K., 1993. Anticonvulsant effect of fluoxetine on focally evoked limbic motor seizures in rats. Epilepsia 34, 381– 384. Rae, J.L., Rich, A., Zamudio, A.C., Candia, O.A., 1995. Effect of Prozac on whole cell ionic currents in lens and corneal epithelia. American Journal of Physiology 269, C250–C256. Speckmann, E.J., Straub, H., Kohling, R., 1993. Contribution of calcium ions to the generation of epileptic activity and antiepileptic calcium antagonism. Neuropsychobiology 27, 122–126. Stauderman, K.A., Gandhi, V.C., Jones, D.J., 1992. Fluoxetineinduced inhibition of synaptosomal [3H]5-HT release: possible Ca2+channel inhibition. Life Sciences 50, 2125–2138. Stefani, A., Spadoni, F., Bernardi, G., 1997. Voltage-activated calcium channels: targets of antiepileptic drug therapy? Epilepsia 38, 959–965. Traynelis, S.F., Dingledine, R., 1988. Potassium-induced spontaneous electrographic seizures in the rat hippocampal slice. Journal of Neurophysiology 59, 259–276. Tytgat, J., Maertens, C., Daenens, P., 1997. Effect of fluoxetine on a neuronal, voltage-dependent potassium channel(Kv1.1). British Journal of Pharmacology 122, 1417–1424. Velasco, A., Alamo, C., Hervas, J., Carvajal, A., 1997. Effects of fluoxetine hydrochloride and fluvoxamine maleate on different preparations of isolated guinea pig and rat organ tissues. General Pharmacology 28, 509–512. Wada, Y., Shiraishi, J., Nakamura, M., Hasegawa, H., 1995. Prolonged but not acute fluoxetine administration produces its inhibitory effect on hippocampal seizures in rats. Psychopharmacology Berlin 118, 305–309. Walden, J., Pockberger, H., Speckmann, E.J., Petsche, H., 1986. Paroxysmal neuronal depolarizations in the rat motorcortex in vivo:intracellular injection of the calcium agonist BAY K 8644. Experimental Brain Research 64, 607–609. Witte, O.W., Speckmann, E.J., Walden, J., 1987. Motor cortical epileptic foci in vivo: actions of a calcium channel blocker on paroxysmal neuronal depolarizations. Electroencephalographic and Clinical Neurophysiology 66, 43–55. Wong, D.T., Bymaster, F.P., Engleman, E.A., 1995. Prozac (fluoxetine, Lilly 110140), the first selective serotonin uptake inhibitor and an antidepressant drug: twenty years since its first publication. Life Sciences 57, 411–441.