Flunarizine inhibits a high-threshold inactivating calcium channel (N-type) in isolated hippocampal neurons

112

Brain Research, 549 (1991) 112-117 © 1991 Elsevier Science Publishers B.V. 0006-8993/91/$03.50 ADONIS 0006899391166047

BRES 16604

Flunarizine inhibits a high-threshold inactivating calcium channel (N-type) in isolated hippocampal neurons Jan Tytgat 1, Petrus J. Pauwels 2, J o h a n Vereecke I and E d w a r d Carmeliet 1 1Laboratory of Physiology, K.U. Leuven, Leuven (Belgium) and 2Department of Biochemical Pharmacology, Janssen Research Foundation, Beerse (Belgium) (Accepted 11 December 1990) Key words: Flunarizine; Ca channel; Hippocampal neurons

The action of flunarizine on the high-threshold inactivating calcium channel (N-type) in hippocampal neurons of the rat was investigated using the whole-cell voltage clamp technique. Flunarizine reduced the currents at all test potentials, without shifting the peak of the current-voltage relationship along the voltage-axis. The drug did not affect the activation curve, but drastically decreased the slope conductance in the linear region of the current-voltage relationship. Block of the current by flunarizine occurred in a use-dependent way. Flunafizine was without effect when applied intraceUulary, and the onset of action, when applied extracellularly, was slow (range of minutes). The Ka for the block by flunarizine obtained after 6 repetitive depolarizations at 0.2 Hz (pulse duration 150 ms) from -90 mV to 0 mV was 0.8 gM. In conclusion, we present eleetrophysiological evidence that flunarizine blocks the high-threshold inactivating Ca channel of hippocampal neurons of the rat. We discuss the possibility that flunarizine might inhibit neuronal transmitter release by means of this effect.

INTRODUCTION It is well known that Ca entry via voltage-gated Ca channels is essential for neurotransmitter release 21. With the voltage clamp technique, several types of Ca channels have been identified in neuronal cells 5"6A°'22. Based on the voltage range of activation of the channels, Carbone and Lux 5'6 have identified low- and high-threshold Ca channels in vertebrate sensory neurons. More recently, Nowycky et al. 22 have shown that high-threshold Ca channels in chick sensory neurons can be classified in inactivating and non-inactivating channels, and they designated the calcium currents as T (low-threshold), N (high-threshold inactivating), and L (high-threshold noninactivating). The discovery of the coexistence of different types of Ca channels has made it possible to investigate the specific role of each of the channel types in neuronal functioning. The T-type current could contribute to rhythmic firing of vertebrate neurons TM, and the N and/or L type current could be involved in the release of neurotransmitters ~6'21'25. Since specific modification of one of the Ca channel types might therefore have important implications for neuronal functioning, a number of agents have been tested on possible selective blocking properties. Several studies reported reduction of the high-threshold inactivating Ca channel by different

agents such as dynorphin A 12, adenosine 2°, acetylcholine 32, w-conotoxin G V I A xg, baclofen 9 (GTP-7-S and GDP-fl-S-dependent), gadolinium 7, protein kinase C activators 8, noradrenaline 17, and [•-Ala,MePhe, Glyol]enkephalin 15. The high-threshold inactivating channel has also been reported to show enhanced inactivation upon the activation of protein kinase C 13, and a noradrenalineand dynorphin A-induced change in the voltage dependency of the channel activation has been reported by Bean 4. In this study, we have investigated the effect of flunarizine, a diphenylpiperazine derivative, on the highthreshold inactivating Ca channel (N-type). Terada et al. 29 have reported that flunarizine acts as a Ca-antagonist in rabbit intestinal smooth muscle cells. We have previously shown that flunarizine blocks L- and T-type Ca channels in cardiac cells 31. Flunarizine exerts a similar effect in rat hypothalamic neurons 1 and rat smooth muscle cells 2. Flunarizine has also been described to bind to Na channels in rat brain 23, and to prevent neurotoxicity induced by prolonged openings of veratridinesensitive Na channels in rat brain neuronal cultures 24. Until now however, no information is available on the effect of flunarizine on the neuronal high-threshold inactivating Ca channel (N-type). In this work, we demonstrate that flunarizine blocks

Correspondence: J. Tytgat, Laboratory of Phsyiology, K.U. Leuvcn, Campus Gasthuisberg, B-3000 Leuven, Belgium. Fax: (32) (16) 21.59.91.

113 the high-threshold inactivating Ca channel in hippocampal neurons in a use-dependent way, and that the Kd for the block by flunarizine obtained after 6 repetitive depolarizations at 0.2 Hz (pulse duration 150 ms) from -90 to 0 mV is 0.8/aM. MATERIALS AND METHODS

Serum-free culture of neurons Cultures of neurons were prepared from the hippocampal formation of 17-day-old rat embryos, as described previously24. Cells were plated on polylysine-coated (0.001%) microscope glass coverslips (Chance propper Ltd. Smetwick, Warley, U.K.) at a density of 1 × 106 cells/coverslip (surface area: 1.13 cm 2) with 0.5 ml DMEM/Ham's F12 (3:1, v/v) plus 10% heat-inactivated horse serum. After 24 h, cultures were switched from serum-containing medium to 1.0 ml serum-free, chemically defined CDM R12

13

7

Measurements The cells were placed in a chamber with a fast perfusion system used to expose the cells to the extracellular solutions and flunarizine within a few seconds. The experiments were performed at room temperature (20-22 °C) with the single electrode whole-cell voltage clamp technique TM using suction pipettes with resistances ranging from 2 to 4 Mr2. The currents were filtered using an 8-pole Bessel filter (3 dB at 500 Hz), and sampled at 2 kHz using a 12-bit ADC (LabMaster TM40, Scientific Solutions, Foster City, U.S.A.). Data acquisition and analysis were controlled by pCLAMP software (Axon Instruments, Foster City, U.S.A.). In order to eliminate K currents, K was replaced in the extra- and intracellular solution by Cs, and 20 mM T E A was added to the intracellular solution. Na currents were eliminated using a Na-free, Tris-containing extracellular solution. With Ba as the extracellular charge carrier, the current evoked from -90 mV consisted of a low-

°i

E

-I

~. _Z-

-2

o

medium 26. Cultures were maintained for one week at 37 °C in an air/CO 2 (95:5) water-saturated atmosphere.

C 0

"E

-4

o

-5

/L i

-50

i

-30

i

i

-10

/ i

i

10

30

feet potential [mV]

E

1 1

~E 0 ¢3 "0 0 N

C

,_, .

0.5

O

E 0 C

50 pA 1 50 ms

-50

-30

- I0

I0

30

test potential [mY] Fig. 1. Effect of flunarizine on high-threshold inactivating Ca channel: current-voltage relationship and steady-state activation. High-threshold inactivating difference currents (see Materials and Methods section) of the same cell in control conditions after 5 (part A) and 15 min (part B) perfnsion, and in the presence of 5/~M flunarizine after 30 rain equilibration of the drug (part C) are shown. Test potentials range from -50 to +30 mV (150-ms pulse duration, 0.2 Hz stimulation frequency). Part D summarizes the peak current-voltage relationship in control conditions (0), and in the presence of 5 ~M flunarizine after respectively 15 ( A ) and 25 rain ( , ) equilibration. Each point represents average current density from at least 5-22 cells. The drug reduces the current over the whole voltage range, without shifting the position of the peak along the voltage axis. Note also the reduction of the linear slope conductance in the region between 0 and +30 mV as induced by 5/~M fiunarizine. Part E shows the lack of effect on steady-state activation of the channel. The half maximial activation potential is -14.8 + 2 mV (n = 17) in control, and -16.7 + 2.3 mV (n = 5) in the presence of 5/zM flunarizine (15 rain). The respective slope parameters are 6.5 _+ 1.2 mV and 6.7 + 0.9 mV.

114 and high-threshold inactivating, and a high-threshold non-inactivating component. The current evoked from -50 mV was defined as the high-threshold non-inactivating current. High- and lowthreshold inactivating currents were separated from the highthreshold non-inactivating current by subtracting the currents evoked at -50 mV from the currents evoked at -90 mV. In order to discriminate then between the high- and low-threshold inactivating current, we have studied the high-threshold inactivating current in conditions where the presence of the low-threshold inactivating current is negligible: i.e. with 1.8 mM Ba o as charge carrier ~, and at test potentials of I>0 mV.

inactivating Ca current. A t the left, high-threshold inactivating Ca currents, m e a s u r e d during a 150-ms depolarization to different test potentials ranging from - 5 0 to +30 mV, are shown in control after 5 min perfusion (part A ) , in control after 15 min perfusion (part B), and after 30 min presence of 5 ~ M flunarizine (part C). It is clear from the current traces that no run-down occurred in control conditions within 15 min, and that the current is completely blocked after 30 min presence of 5 p M of the drug. Part D of the figure summarizes the p e a k currentvoltage relationship in control conditions, and in the presence of 5 p M flunarizine (after 15 and 25 min equilibration). Each point represents the average current density from at least 5-22 cells. The curves represent a fit of the data points with the product of a linear fully activated current-voltage relation times a Boltzmann function representing steady-state activation. 5 p M flunarizine r e d u c e d the current over the whole voltage range without shifting the position of the p e a k , situated near - 5 mV, along the voltage axis (1.8 m M Ba o, 150 ms pulse duration, 0.2 Hz stimulation rate). The drug decreased the slope conductance in the linear region

Solutions during electricalmeasurements The cells were superfused with an extracellular solution (= bath solution) containing (in mM): 137.6 Tris, 1 MgCI2, 1.8 BaCI2, 5 glucose, 20 CsCI, and titrated with HCI to pH 7.4. The intracellular solution (= pipette solution) contained (in raM): 125 CsCI, 5 MgATP, 15 EGTA, 20 TEA-CI, 10 Hepes, and brought to pH 7.2 with CsOH. Flunarizine (kindly provided by Janssen Pharmaceutica, Beerse, Belgium) was added to the extracellular solution. RESULTS

Effect of flunarizine on the high-threshold inactivating Ca channel: current-voltage relationship and steady-state activation Fig. 1 shows a typical result of the effect of flunarizine on the current-voltage relationship of the high-threshold

A

e

|

b-a=

100pAL___

b-a=A b-a

•

25 ms

:•

control

D

5 p,M flu

10 p,M flu

VV

= -100 "E

V

&& A

o

00000 -200

0

10 20

325

345

650

670

time [S]

Fig. 2. Flunarizine block of high-threshold inactivating Ca channel shows use-dependency. A train of 6 depolarizations (150 ms pulse duration, 0.2 Hz stimulation frequency) from -90 mV (b) and -50 mV (a) to 0 mV is applied in control conditions (A), and in the presence of 5 (B) and 10/~M flunarizine (flu) (C). The development of the use-dependent block on the difference current (b-a) as a function of time is shown in part D. It is clear that repetitive stimulation of the channel in the presence of flunarizine induces a faster development of block of the channel than periods of rest without stimulation.

115 depolarizations from -90 mV promoted high-threshold inactivating Ca channel block by flunarizine (see parts B b and Cb). When the holding potential was set at -50 mV, a strong tonic block was observed (see parts B a and Ca). The onset of block by flunarizine, after the drug was applied extracellularly, was slow (range of minutes). Flunarizine applied intracellularly (data not shown), was without effect on the currents.

100

(4) 75

50

(lO) ~

o=

]

25

0.01

0.1

1

10

flunarlzlne [/zM]

Fig. 3. Dose-response curve of the effect of flunarizine on the amplitude of high-threshold inactivating current. The remaining current (as difference current in percent), plotted as a function of the concentration of the drug, is determined as the current after a train of 6 repetitive, 150 ms depolarizations to 0 mV at 0.2 Hz. In total, 47 cells are tested. The smooth curve represents a fit of the data points with a Kd of 0.8/~M and a Hill coefficient of 0.7. between 0 and +30 mV to 22% of the value in control. Part E of the figure shows the effect of flunarizine on steady-state activation. The curves were obtained by dividing the currents from Fig. 1D in the range between -50 and +30 mV by the fully activated current, which was estimated by extrapolation of a linear regression through the currents between 0 and +30 mV. In controls (n = 17), the potential of half maximal activation was -14.8 _+ 2 mV, and the slope parameter 6.5 _+ 1.2 mV. After 15 min equilibration of the cell in the presence of 5 /~M flunarizine (n = 5), the steady-state activation curve was characterized by a half maximal potential of activation of -16.7 + 2.3 mV, and a slope parameter of 6.7 + 0.9 mV. These results demonstrate that flunarizine does not significantly affect the steady-state activation curve (Student's t-test, P < 0.05).

Block of the high-threshold inactivating Ca channel by flunarizine is use-dependent To study possible use-dependent block of the highthreshold inactivating Ca channel by flunarizine, a train of 6 repetitive 150-ms step depolarizations to 0 mV from a holding potential o f - 9 0 and -50 mV was applied at 0.2 Hz. This protocol was applied in control conditions (Fig. 2, part A), as well as in the presence of 5/~M flunarizine (part B), and on the same cell in the presence of 10 ~M flunarizine (part C), each after a 5-min period without stimulation. Part D summarizes the development of block as a function of time. In order to measure the use-dependent block, the 6 currents evoked during a train of pulses from holding potential -50 mV were subtracted from the corresponding 6 currents evoked during a train of pulses from holding potential -90 mV. It is clear that repetitive

Dose-response curve of flunarizine block on the highthreshold inactivating Ca channel Figure 3 shows the dose-response curve of block by flunarizine on the high-threshold inactivating Ca channel. The curve shows the current left after a train of 6 repetitive, 150-ms depolarizations to 0 mV at 0.2 Hz, after a 15-min period at -90 mV in the presence of flunarizine. The current was expressed relatively to the value in control conditions. In control, the application of 6 depolarizations reduced the current only to a very small amount (2.9 + 1.3%, n -- 47). The smooth curve represents a fit of the data points with a K o of 0.8~M and a Hill coefficient of 0.7. DISCUSSION

Our experiments show that flunarizine blocks the high-threshold inactivating Ca channel (N-type) in hippocampal neurons. This channel was measured in conditions in which Ca o was replaced by 1.8 mM Ba o (which reduces to a large extent the low-threshold inactivating current), and by subtracting the current evoked from a holding potential o f - 5 0 mV (to eliminate the highthreshold non-inactivating current), since it is known that the high-threshold inactivating current is for the greatest part inactivated at -50 mV 11'28. However, the inactivation range of the high-threshold inactivating current is spread over a broad potential range and its inactivation might not be complete at -50 m V TM. AS a consequence, subtracting the current evoked from a holding potential o f - 5 0 mV, to eliminate the high-threshold non-inactivating Ca current, might have resulted in an underestimation of the high-threshold inactivating current. Nevertheless, the following observations indicate that the subtraction method is a reliable tool to study the high-threshold inactivating Ca current, and isolates a Ca current type which resembles the 'N-type' Ca current, rather than the 'T-type' (low-threshold inactivating), or 'L-type' (high-threshold non-inactivating) Ca current as observed in chick D R G cells by Nowycky et alZ2: (1) Difference currents obtained in experiments performed with 10.8 instead of 1.8 mM Ba o (data not shown), gave rise to clearly distinct peaks in the current-voltage relationship (around -30 and +10 mV). Both peaks

116 corresponded to an inactivating current, which suggest the existence of two distinct types of inactivating channels corresponding to the T- and N-type Ca channel, as observed in the same preparation by Takahashi et al. 28. In our experiments performed with 1.8 mM Bao, the current-voltage relationship showed only one peak around -5 mV (see Fig. 1D). Since the current density of the N-type Ca current in chick sensory neurons was about 8 times larger than the T-type 22, and since Akaike et al.1 demonstrated that activity of the low-threshold Ca channel in rat hypothalamic neurons is virtually absent when Ba o is as low as 1.8 mM, the difference current in our experiment with 1.8 mM Ba o was most likely the N-type. (2) The activation range of the high-threshold inactivating current in our experiments was similar to the one of the N-type Ca channel (positive to -50 mV) and differed from the activation ranges of other Ca conductances in rat hippocampal CA1 pyramidal cells as described by Takahashi et al. 28. (3) The current obtained with our subtraction procedure had similar characteristics (activation and inactivation range, time constant of inactivation) as the current most sensitive to 1-5 pM w-conotoxin G V I A (some data not shown). Aosaki and Kasai 3 have shown on chick sensory neurons that 5/~M to-conotoxin G V I A irreversibly eliminated only the small-conductance high-voltage-activated Ca channel, while the large-conductance channel was either unaffected or only transiently blocked. (4) In some cells where a non-inactivating current (evoked from -50 mV) was absent or minimal, the block by flunarizine of the high-threshold inactivating current was similar to the block in the other cells. In summary, it is feasible to designate the difference currents described in this study as N-type Ca currents similar to the ones described by Nowycky et al. z2 in chick D R G cells and by Takahashi et al. 2s in hippocampal neurons. Electrophysiological data on single cells demonstrate that flunarizine possesses Ca-antagonistic properties with voltage- and use-dependent characteristics. Terada et al. 29 have shown that flunarizine blocked the L-type channel in fragmented smooth muscle cells of the rabbit small intestine with a K d of 1.4 p M (holding potential -60 mV), that the drug shifted the inactivation curve to more negative potentials, that the relative suppression of the current-voltage relationship was indentical at all potentials, and that the drug had a slow onset of action. In 1988, Tytgat et al. 31 have demonstrated that flunarizine was the first organic Ca antagonist with L- and T-type Ca channel blocking properties on heart (1-10/~M). Similar results were obtained in rat hypothalamic neurons and rat smooth muscle cells by Akaike et al. 1'2, and a higher affinity of block than on heart muscle was found (Ko

0.09-3.2/tM, depending on the experimental conditions). In our study, we demonstrate that flunarizine blocks the high-threshold inactivating Ca channel with a K d of 0.8 pM for 6 repetitive 150-ms depolarizations at 0.2 Hz from -90 to 0 mV. Since the Kd-value is in the same range as the value reported for the block of the T-type channel in neuronal cells 1, it can therefore be assumed that T-type current and high-threshold inactivating current (N-type) are about equally sensitive to block by flunarizine (although comparison of K d values is often difficult, since they are strongly dependent on experimental conditions such as holding potential, frequency of stimulation, and external concentration of charge carrier). The action of flunarizine is not restricted to Ca channels. Evidence for inhibitory effects on neuronal Na channels was given by Pauwels et al. 23. These authors have reported that flunarizine competed with [3H]batrachotoxinin A20-a-benzoate binding to Na channels and veratridine-induced Na uptake in rat brain synaptosomal preparations (K d values 0.3 and 0.2 #M, respectively). In 1989, Pauwels et al. z4 have also observed that I p M flunarizine prevented Ca 2+-dependent, veratridineinduced, neuronal cell loss in primary rat brain neuronal cultures. Also in cardiac muscle, we have observed an inhibitory action of flunarizine on the Na channel (unpublished observations of two electrode whole cell clamp experiments in conditions where the channel was partly inactivated). The functional role of different types of Ca channels in neurons is still subject to speculation 21. More recent data by Hirning et al. 16 demonstrate a role of N-type Ca channels in neurotransmitter release. These authors showed that block of the L-type Ca channel by nitrendipine was not sufficient to inhibit depolarization-induced release of neurotransmitter in sympathetic ganglion cells, but that block of both L- and N-type channel by to-conotoxin G V I A prevented neurotransmitter release. Silverstein et al. 27 have demonstrated that flunarizine limits striatal dopamine release induced by hypoxiaischemia. Recently, Tertian et al. 3° have reported that flunarizine was a potent inhibitor of depolarizationinduced Ca mobilization and the release of dynorphin A(1-8)-like immunoreactivity (Dyn-LI) in rat hippocampal mossy fiber synaptosomes. Tertian et al. 3° have also shown that the release of Dyn-Ll was relatively insensitive to dihydropyridines, verapamii, and diltiazem (which are L-type blockers), and to amiloride and phenytoin (which are T-type blockers). They concluded that presynaptic N-type Ca channels make a substantial contribution to the Ca influx required for the exocytosis of Dyn-LI from hippocampal mossy fiber nerve endings. Based on our electrophysiological experiments, inhibition of transmitter release in hippocampal neurons might

117 N-type Ca channel is solely responsible for the flunari-

be explained by inhibition of a high-threshold inactivating Ca current (N-type) by flunarizine. Yet, this interpretation should be approached with caution since our studies

zine-induced decrease in transmitter release in vivo.

were c a r d e d out on an 'artificial' system (cultured neurons, defined m e d i u m , perfusion solutions). Therefore, it is difficult to demonstrate that the block of the

Acknowledgements. J.T. was supported by a grant of the National Fund for Scientific Research (Belgium) of which he is a Research Assistant. We thank Mrs. L. Heremans, Mrs. D. Hermans, Mr. M. Coenen, and Mr. R. Verbist for technical support.

REFERENCES 1 Akaike, N., Kostyuk, P.G. and Osipchuk, Y.V., Dihydropyridine-sensitive calcium channels in isolated rat hypothalamic neurones, J. Physiol., 412 (1989) 181-195. 2 Akaike, N., Kanaide, H., Kuga, T., Nakamura, M., Saoshima, J.I. and Tomoike, H., Low-voltage-activated calcium current in rat aorta smooth muscle cells in primary culture, J. Physiol., 416 (1989) 141-160. 3 Aosaki, T. and Kasai, H., Characterisation of two kinds of high-voltage-activated Ca-channel currents in chick sensory neurons, Pflag. Arch. Eur. J. Physiol., 414 (1989) 150-156. 4 Bean, B.P., Neurotransmitter inhibition of neuronal calcium currents by changes in channel voltage dependence, Nature, 340 (1989) 153-156. 5 Carbone, E. and Lux, H.D., A low voltage-activated calcium conductance in embryonic chick sensory neurones, Biophys. J., 46 (1984) 413-418. 6 Carbone, E. and Lux, H.D., A low voltage-activated, fully inactivating Ca channel in vertebrate sensory neurones, Nature, 310 (1984) 501-503. 7 Docherty, R.J., Gadolinium selectively blocks a component of calcium current in rodent neuroblastoma x glioma hybrid (NG108-15) cells, J. Physiol., 398 (1988) 33-47. 8 Doerner, D., Pitler, T.A. and Alger, B.E., Protein kinase C activators block specific calcium and potassium current components in isolated hippocampal neurons, J. Neurosci., 8 (1988) 4069-4078. 9 Dolphin, A.C. and Scott, R.H., Calcium channel currents and their inhibtion by (-)baclofen in rat sensory neurons: modulation by guanine nucleotides, J. Physiol., 386 (1987) 1-17. 10 Fedulova, S.A., Kostyuk, EG. and Veselovsky, N.S., Two types of calcium channels in the somatic membrane of new-born rat dorsal root ganglion neurons, J. Physiol., 359 (1985) 431-446. 11 Fox, A.P. Nowycky, M.C. and Tsien, R.W., Kinetic and pharmacologic properties distinguishing three types of calcium currents in chick sensory neurones, Jr. Physiol., 394 (1987) 149-172. 12 Gross, R.A. and MacDonald, R.L., Dynorphin A selectively reduces a large transient (N-type) calcium current of mouse dorsal root ganglion neurons in cell culture, Proc. Natl. Acad. Sci. U.S.A., 84 (1987) 5469-5473. 13 Gross, R.A. and MacDonald, R.L., Activators of protein kinase C selectively enhance inactivation of a calcium current component of cultured sensory neurons in a pertussis toxin-sensitive manner, J. Neurophysiol., 61 (1989) 1259-1269. 14 Hamill, O.P., Marty, A., Neher, E., Sakmann, B. and Sigworth, E, Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches, Pflag. Arch. Eur. J. Physiol., 391 (1981) 85-100. 15 Henderson, G. and Seward, E.E, Inhibitionof an N-like calcium channel current by/~-opioid receptor activation in the human neuroblastoma cell line SH-SY5Y, J. Physiol., 422 (1990) 19E 16 Hirning, L.D., Fox, A.P., McCleskey, E.W., Olivera, B.M., Thayer, S.A., Miller, R.J. and Tsien, R.W., Dominant role of N-type Ca2+ channels in evoked release of norepinephrine from sympathetic neurons, Science, 239 (1988) 57-61. 17 Lipscombe, D., Kongsamut, S. and Tsien, R.W., a-Adrenergic inhibition of sympathetic neurotransmitter release mediated by

modulation of N-type calcium-channel gating, Nature, 340 (1989) 639-642. 18 Llinas, R. and Yarom, Y., Properties and distribution of ionic conductances generating electroresponsiveness of mammalian inferior olivary neurones in vitro, J. Physiol., 315 (1981) 569-584. 19 McCleskey, E.W., Fox, A.P., Feidman, D.H., Cruz, L.J., Olivera, B.M., Tsien, R.W. and Yoshikami, D., to-Conotoxin: direct and persistent blockade of specific types of calcium channels in neurons but not muscle, Proc. Natl. Acad. Sci. U.S.A., 84 (1987) 4327-4331. 20 Madison, D.V., Fox, A.P. and Tsien, R.W., Adenosine reduces an activating component of calcium current in hippocampal CA3 neurons, Biophys. J., 51 (1987) 30a. 21 Miller, R.J., Multiple calcium channels and neuronal function, Science, 235 (1987) 46-52. 22 Nowycky, M.C., Fox, A.P. and Tsien, R.W., Three types of neuronal calcium channel with different calcium agonist sensitivity, Nature, 316 (1985) 440-443. 23 Pauwels, P.J., Leysen, J.E. and Laduron, P.M., [3H]Batrachotoxinin A 20-a-benzoate binding to sodium channels in rat brain: characterization and pharmacological significance, Eur. J. Pharmacol., 124 (1986) 291-298. 24 Pauwels, P.J., Van Assouw, H.P., Leysen, J.E. and Janssen, P.A.J., Ca÷+-mediated neuronal death in rat brain neuronal cultures by veratridine: protection by flunarizine, Mol. Pharmacol., 36 (1989) 525-531. 25 Perney, T.M., Hirning, L.D., Leeman, S.E. and Miller, R.J., Multiple calcium channels mediate neurotransmitter release from peripheral neurons, Proc. Natl. Acad. Sci. U.S.A., 83 (1986) 6656-6659. 26 Romijn, H.J., Van Huizen, E and Wolters, P.S., Towards an improved serum-free, chemically defined medium for long-term culturing of cerebral cortex tissue, Neurosci. Biobehav. Rev., 8 (1984) 301-334. 27 Silverstein, E, Buchanan, K. and Johnston, M.V., Flunarizine limits striatal dopamine release induced by hypoxia-ischemia, Soc. Neurosci., Abstr., 10 (1984) 66. 28 Takahashi, K., Wakamori, M. and Akaike, N., Hippocampai CA 1 pyramidal cells of rats have four voltage-dependent calcium conductances, Neurosci. Lett., 104 (1989) 229-234. 29 Terada, K., Ohya, Y., Kitamura, K. and Kuriyama, H., Actions of flunarizine, a Ca ++ antagonist, on ionic currents in fragmented smooth muscle cells of the rabbit small intestine, J. Pharmacol. Exp. Ther., 240 (1987) 978-983. 30 Terrian, D.M., Damron, D.S., Dorman, R.V. and Gannon, R.L., Effects of calcium antagonists on the evoked release of dynorphin A(1-8) and availability of intraterminal calcium in rat hippocampal mossy fiber synaptosomes, Neurosci. Lett., 106 (1989) 322-327. 31 Tytgat, J., Vereecke, J. and Carmeliet, E., Differential effects of verapamil and flunarizine on cardiac L-type and T-type Ca channels, Naunyn-Schmiedeberg' s Arch. Pharmacol., 337 (1988) 690-692. 32 Wanke, E., Ferroni, A., Malgaroli, A., Ambrosini, A., Pozzan, T. and Meldolesi, J., Activation of a muscarinic receptor selectively inhibits a rapidly inactivated Ca2+ current in rat sympathetic neurons, Proc. Natl. Acad. Sci. U.S.A., 84 (1987) 4314-4317.