Gabapentin inhibits calcium currents in isolated rat brain neurons

Neuropharmacology 37 (1998) 83 – 91

Gabapentin inhibits calcium currents in isolated rat brain neurons Alessandro Stefani a,b,*, Francesca Spadoni a,b, Giorgio Bernardi a,b a

b

IRCCS Ospedale S. Lucia, Via Ardeatina, 306 Rome, Italy Uni6ersita` di Roma Tor Vergata, Clinica Neurologica c/o Dip. Sanita` Pubblica, Via di Tor Vergata, 13500135 Rome, Italy Accepted 9 October 1997

Abstract Gabapentin (1(aminomethyl) cyclohexane acetic acid; GBP) is a recently developed anticonvulsant, for which the mechanism of action remains quite elusive. Besides its possible interaction with glutamate synthesis and/or GABA release, in cerebral membranes gabapentin has been shown to bind directly to the a2d subunit of the calcium channel. Therefore, we have tested the possibility that gabapentin affects high threshold calcium currents in central neurons. Calcium currents were recorded in whole-cell patch-clamp mode in neurons isolated from neocortex, striatum and external globus pallidus of the adult rat brain. A large inhibition of calcium currents by gabapentin was observed in pyramidal neocortical cells (up to 34%). Significantly, the gabapentin-mediated inhibition of calcium currents saturated at particularly low concentrations (around 10 mM), at least in neocortical neurons (IC50 about 4 mM). A less significant inhibition was seen in medium spiny neurons isolated from striatum (−12.4%) and in large globus pallidus cells ( −10.4%). In all these areas, however, the GBP-induced block was fast and largely voltage-independent. Dihydropyridines (nimodipine, nifedipine) prevented the gabapentin response. v-conotoxin GVIA and v-conotoxin MVIIC, known to interfere with the currents driven by a1b and a1a calcium channels, did not prevent but partially reduced the response. These findings imply that voltage-gated calcium channels, predominately the L-type channel, are a direct target of gabapentin and may support its use in different clinical conditions, in which intracellular calcium accumulation plays a central role in neuronal excitability and the development of cellular damage. © 1998 Elsevier Science Ltd. All rights reserved. Keywords: Anticonvulsants; Dihydropyridines; Calcium currents; Neuroprotective agents; Gabapentin

1. Introduction Several novel antiepileptic agents, such as lamotrigine, riluzole, topiramate and gabapentin, have been introduced successfully in the last few years (Macdonald and Kelly, 1994; Meldrum 1996; Macdonald and Greenfield, 1997). Aside from their effectiveness as anticonvulsants, these drugs are currently being investigated as promising neuroprotectants (Stefani et al., 1997a). Gabapentin (1 (aminomethyl) cyclohexane acetic acid; GBP), in particular, is used as add-on therapy for partial and secondary generalized seizures (Leiderman, 1994; Beydoun et al., 1995) but it is also utilized in a rather broad range of diseases, including neurodegenerative pathologies, restless legs syndrome, chronic neuralgia and others (Welty et al., 1995; * Corresponding author. Tel.: + 39 6 72596137; fax: + 39 6 72596006; e-mail: ‘‘[email protected]’’. 0028-3908/98/$19.00 © 1998 Elsevier Science Ltd. All rights reserved. PII: S0028-3908(97)00189-5

Cochran, 1996; Gurney et al., 1996; Kanthasamy et al., 1996; Mellick and Mellick, 1996; Patel and Naritoku, 1996; Segal and Rordorf, 1996). Yet, its precise mechanism of action is far from being established. When tested on different enzymes of the metabolic pathways of glutamate and GABA in 6i6o, GBP was demonstrated to inhibit potently the branched-chain amino acid aminotransferase (BCAA-T) (Goldlust et al., 1995), thus reducing the endogenous synthesis of glutamate. In addition, in patients taking GBP, GABA levels were shown to increase (and in a dose-dependent fashion, Petroff et al., 1996), supporting a potential GBP-mediated facilitation of inhibitory transmission (Kocsis and Honmou, 1994; Honmou et al., 1995). GBP was also capable, after prolonged preincubation, to limit the sodium-driven repetitive firing in mammalian cell cultures (Wamil and McLean, 1994). From pig cerebral cortex, however, a GBP-binding protein has recently been isolated, of which the N-terminal

84

A. Stefani et al. / Neuropharmacology 37 (1998) 83–91

sequence matched with the a2d subunit of voltagegated calcium (Ca2 + ) channels (VGCC) (Gee et al., 1996). The brain a2d subunit, an alternatively spliced form of the skeletal muscle subunit (Witcher et al., 1993; Isom et al., 1994), derives from a single gene (Ellis et al., 1988) and through functional coexpression with the different a1 subunits which form the Ca2 + channel pores (Mikami et al., 1989; Williams et al., 1992; Brust et al., 1993), contributes effectively to the stimulation of Ca2 + current amplitude (Catterall, 1995; Gurnett et al., 1996). These findings might imply that VGCC are also one of the critical targets at which GBP exerts it actions. At present, this possibility has not been sufficiently addressed. The efficacy of several AEDs may depend, at least in part, on their ability to reduce calcium (Ca2 + ) conductances (Stefani et al., 1997a). The findings obtained by our group in both isolated neurons and brain slice preparations has provided strong evidence of the inhibition of VGCC by lamotrigine (LTG), riluzole, oxcarbazepine (OCBZ) and felbamate (FBM) (Stefani et al., 1995; Calabresi et al., 1995, 1996; Siniscalchi et al., 1996; Stefani et al., 1996a,b). Since the accumulation of intracellular Ca2 + is a crucial step in the spread of epileptic discharges as well as in the sequence of events leading to cell damage and death (Siesjo and Bengtsson, 1989; Tymianski and Tator, 1996), we have suggested that the AED-mediated reduction of voltage-dependent Ca2 + fluxes underlies their potential usefulness as neuroprotective agents (Stefani et al., 1997a,b). In particular, the reduction of the Ca2 + fluxes through channels which are known to govern transmitter release at axon terminals (N and P/Q type channels) (Takahashi and Momiyama, 1993) caused LTG and OCBZ to have a powerful modulatory effect on glutamate-mediated synaptic potentials (Calabresi et al., 1995, 1996). The inhibition of predominantly L-type channels (mainly distributed in the somatodendritic regions) by FBM should be reflected as an impact on cellular integrative properties and excitability (Stefani et al., 1996b, 1997a). The physiological impact of drugs interfering with regulatory subunits of the Ca2 + channels, such as the a2d subunit, is less predictable. Using molecular investigations, however, it has been clearly documented how procedures which manipulate the a2d subunit (such as N-glycosylation or cleavage into disulfide-linked a2 and d subunits; Gurnett et al., 1996) may reduce dramatically the stimulated Ca2 + current, even in the absence of consistent changes in the voltage-dependence of its activation. For many AEDs, the mechanisms responsible for anticonvulsant activity may be a combination of effects at different receptors or channels (Macdonald and Greenfield, 1997). Phenytoin is known to decrease Ca2 + currents, but only at supratherapeutic concentrations, well above those required to inhibit action poten-

tial discharge (McLean and Macdonald, 1983); whereas low mM concentrations of LTG have been shown to decrease both sodium and Ca2 + -dependent events (Stefani et al., 1996a, 1997b). For FBM, the block of the inactivated sodium channel and the reduction of NMDA-mediated transmission occur in the 50–300 mM range (Pisani et al., 1996); yet, the saturating dose for the inhibition of L-type Ca2 + currents by FBM is close to 500 nM (Stefani et al., 1996b). These examples emphasize the opportunity to characterize the different pharmacological aspects of these AED’s which have been recently introduced. The aim of our study was to assess the putative effect of GBP on high-voltage-activated (HVA) Ca2 + currents in rat central neurons. Since GBP is used in several disease states, presumed to involve both cortical and subcortical regions, GBP-mediated responses were evaluated in different structures, namely cortical, striatal and pallidal neurons.

2. Methods Neocortical, striatal and pallidal neurons were dissociated from 60 male Wistar rats aged 1–2 months. Briefly, as previously reported (Stefani et al., 1996b, 1997b) either neostriatum or the surrounding neocortex or external GP was dissected under stereomicroscope from coronal slices 350–400 mm thick. Slices were incubated in a Hepes-buffered Hank’s balanced salt solution (HBSS), bubbled with 100% O2 and warmed at 35°C. From 30 to 60 min later, one slice (two microslices for GP) was transferred in HBSS media to which 1.5 mg/ml protease XIV had been added. After 30–40 min of enzymatic treatment, the tissue was rinsed in HBSS and mechanically triturated. The cell suspension was then placed in a Petri dish mounted on the stage of an inverted microscope. Cells were allowed to settle for 10–12 min. Neocortical, striatal and pallidal neurons were chosen for recordings if presumed to be, respectively, medium to large pyramidal-shaped cells (usually with a typical apical process spared by the brief enzyme incubation time), medium-spiny neurons (MSNs; usually bipolar, major axis around 15 mm) and large pallidal cells. Patch-clamp recordings in the whole-cell configuration were carried out using pipettes (Corning 7052) pulled at a Flaming-Brown and fire-polished just prior to use. Pipette resistance ranged from 3 to 8 Mohms when filled by the internal solution consisting of (in mM): N-methyl-D-glucamine 185, Hepes 40, EGTA 11, Mg 4, phosphocreatine 20, ATP 2–4, GTP 0–0.2, leupeptin 0.2; pH was adjusted to 7.3 with phosphoric acid; the osmolarity was 275–280 mOsm/l. After obtaining cell access, the neuron was usually bathed in a medium composed of (in mM): TEACl 165, BaCl2

A. Stefani et al. / Neuropharmacology 37 (1998) 83–91

85

Fig. 1. Voltage-gated Ba2 + currents in neocortical neurons and GBP-mediated modulation. Ba2 + currents are evoked by progressively more depolarized voltage steps (from − 55 to + 55 mV, 10 mV interval step) in control (A) and under 10 mM GBP (B). Holding potential was − 60 mV. (C) Current-voltage relationship, from data in (A) and (B) (data points indicated by circles). In (D) plot of the inhibition of Ba2 + currents versus test pulse potential derived from IV curve before and during 10 mM GBP.

2.5 – 5, CsCl2 5, Hepes 10; pH was adjusted to 7.4 and the osmolarity to 300 – 305 with glucose. Only in a subset of recordings (n =20) NaCl 135 – 140 mM was substituted for TEA as the main cation (in this situation TTX 0.001 mM was added). Control as well as drug solutions were applied with a linear array of six, gravity-fed capillaries positioned within 500 mm of the patched neuron. This system allowed drugs to be applied and washed within s at well-defined concentrations. Recordings were made with an Axopatch 1D at room temperature (21 – 22°C). Series resistance compensation (70–80%) was routinely employed. Data were low-pass filtered (corner frequency=5 kHz). For data acquisition and analysis pClamp 5.51 running on a Pentium PC was used. Ba2 + currents were studied with voltage steps and ramps. Ramp speed (0.3 – 0.6 mV/ms) was chosen to maximize the agreement between the current/voltage relationship obtained with this method and that derived from short (30 ms) step depolarizations. Values given in the text and in the figures are mean 9SEM of changes in the respective cell populations. Student’s t-test (for paired and unpaired observations) was used to compare the means. All compounds were obtained by Sigma. Nifedipine-containing solutions were light-protected. GBP was a kind gift of Parke-Davis.

3. Results In this study, we have investigated the GBP effect on VGCC by: (i) characterizing the GBP-mediated inhibition of HVA Ca2 + currents isolated in neurons obtained from cortex; (ii) comparing the modulation produced by GBP in the neocortex with the inhibitory response caused in the striatum and GP; (iii) identifying, by utilizing selective channel blockers, the Ca2 + channel types targeted by GBP. As previously described (Sayer et al., 1993; Lorenzon and Foehring, 1995; Stefani et al., 1996a,b), in isolated cells from the adult rat brain, HVA Ca2 + currents may be activated either by depolarizing voltage steps (Figs. 1 and 2) or by a voltage ramp (0.3 mV/ms; Fig. 3A). From a holding potential of − 50/− 60 mV, the pattern of depolarization-activated currents is dominated by high-threshold, sustained, non-inactivating Ba2 + conductances, which peak around − 10 mV and are abolished by 100 mM Cadmium (Cd2 + ) (data not shown). Even when present, the contribution of lowvoltage activated Ca2 + currents is negligible, because of the holding potential or the tissue development. GBP was tested on cortical pyramidal cells. GBP decreased step-activated calcium currents in almost all the patched pyramidal cortical cells (36 out of 38); a

86

A. Stefani et al. / Neuropharmacology 37 (1998) 83–91

Fig. 2. Test-activated Ca2 + currents (at − 10 mV) preceeded or not by depolarizing prepulses at + 80 mV. The facilitation protocol produces an enhancement of HVA Ca2 + currents by about 18% (A versus a in A). Under GBP (B), both current traces are reduced (B, b). In (C–D), the same current traces as in (A–B), to show that the GBP response is rather similar; in fact, the inhibition is 26.52% for standard conductance (a–b in C) and 26.24% for facilitated currents (A–B in D).

typical response is depicted in Fig. 1A – B; the inhibition was neither paralleled by a clear slowing of the activation nor was it strongly voltage-dependent, as illustrated by the current-voltage relationship and furthermore, by the percentage inhibition at each voltage (Fig. 1C–D). These findings seem to suggest that the GBP-mediated inhibition differs from classical modulators known to interfere with N- and P-type channels (such as agonists at metabotropic glutamate receptors, Choi and Lovinger 1996; Stefani et al., 1997c), in that it probably does not interact with the G-protein regulated voltage sensor of the channel. The GBP response, however, developed fast and did not show any desensitization after repeated applications (data not shown).Additional analysis of the GBP-induced inhibition was carried out by evaluating the effects of GBP on standard and facilitated currents (after the interposition of a pre-depolarizing step to + 80/100 mV; Fig. 2A – B). The percentage inhibition by saturating concentrations of GBP (30 mM) did not differ significantly (Fig. 2C – D). Analogous findings were observed in five other neurons. The inhibition of ‘standard’ and ‘‘facilitated currents was 33.26 (94.06, n = 10) and 32.24 (9 4.84, n = 10), respectively.

GBP inhibition of HVA Ca2 + currents, studied at a broad range of concentrations (from nM up to 100 mM) was dose-dependent and fully reversible (Fig. 3). The GBP-mediated reduction of VGCC saturated at 10 mM (Fig. 3B); a threshold response was observed with 0.5–1 mM (Fig. 3B). Maximal inhibition was close to 34% (34.329 4.42, n=12). The IC50 was calculated as 3.65 mM (Fig. 3 and Table 1). We examined whether the large GBP-mediated inhibition of HVA Ca2 + currents was selective for cortical neurons or was also present in other areas, namely striatum and GP. The GBP-induced reduction of HVA Ca2 + currents, although consistently observed in both striatal (18 out of 20 recordings) and pallidal (23 out of 25 recordings) neurons, was rather small. The maximal response averaged slightly more than 12 and 10%, respectively (Table 1).It is difficult to know whether such a striking discrepancy is due to a different expression of GBP binding sites in cells isolated from different brain regions. GBP responses, however, showed a quite similar IC50 in all the areas tested (Table 1). Alternatively, the different potency of GBP in the cortex, compared with the GP and striatum, might be due to a different functional expression of the Ca2 +

A. Stefani et al. / Neuropharmacology 37 (1998) 83–91

87

Fig. 3. (A) Ramp-activated (0.5 ms/mV) whole-cell currents from − 70 mV under control conditions and three sequential concentrations of GBP. Note that the response saturates at 10 mM. (B) Dose-response of GBP-mediated inhibition of cortical Ca2 + currents. The graph illustrates the percent inhibition of HVA Ca2 + currents by different concentrations of GBP. Each point represents the mean of at least six experiments. Fitting was done with the following equation: m3(m0 ‚ m1/m0 ‚ m1 +m2 ‚ m1). IC50 =3.65 mM.

channels targeted by GBP in each structure. In order to identity the Ca2 + channel which is inhibited by GBP we have utilised the usual pharmacological agents which are selective blockers for different Ca2 + channel subtypes: the dihydropyridines (DHP) nimodipine and nifedipine, v-CgTx and v-CgMVIIC for respectively L- and N-, and P/Q currents. Both DHPs, at mM concentrations, are known to inhibit L-type Ca2 + channel currents in mammalian central neurons. DHPsensitive L channels, in the neocortex, account for about 35–39% of the available pool of VGCC from a depolarized holding potential (Lorenzon and Foehring, 1995; Stefani et al., 1995). In striatum and pallidus, the prevalence of L-type Ca2 + channels is smaller, although they still represent about a quarter of the activated current in whole-cell mode (Stefani et al., 1994; Surmeier et al., 1994). 5 mM nimodipine (Fig. 5; n= 7) or nifedipine (n =6) fully prevented the GBP-mediated reduction of VGCC in the neocortex (Fig. 4A), but analogous findings were observed in striatal (n= 5) and pallidus (n=5) cells (Fig. 4b). In every cellular phenotype we studied, the residual inhibition by GBP, after DHP, was negligible (Fig. 4B). Therefore, DHPsensitive L-channels are directly inhibited by GBP, in a Table 1 Comparative data of the GBP-induced inhibition of Ca2+ currents in the cortex, striatum and GP Structure

Maximum inhibition (%)

IC50 (mM)

Cortex Striatum Globus pallidus

34.3 10.4 12.4

3.65 13.84 4.52

Each value is an average of five experiments.

similar manner to that shown for felbamate (Stefani et al., 1996b). As a consequence, the different predominance of the L-type channel in the brain regions examined may explain, albeit partially, the different potency of GBP. If the L-type Ca2 + channel was the exclusive target of GBP, however, then it might be expected that the GBP-induced reduction of Ca2 + currents would not be affected at all by the pharmacological block of other channel subtypes. Instead, the GBP inhibition was not fully resistant to v-CgTx VIA and v-CgTxMVIIC. As revealed in Fig. 5A (and illustrated in the histogram of Fig. 5B), part of the GBP response in cortical cells was antagonised by either 2 mM v-CgTx GVIA (n=4) and 2 mM v-CgTx MVIIC (n= 4) (the average GBP inhibition was reduced from − 33% to about 20% by each toxin, Fig. 5B).

4. Discussion Biochemical studies have proposed that GBP binds to the a2d subunit of VGCC (Taylor et al., 1993; Gee et al., 1996). Recently, a renewed interest has focused on the roles played by the auxiliary a2d subunit (Isom et al., 1994; Catterall, 1995); in particular, it was shown unequivocally that its coexpression (and coassembly with the a1 subunit) is required for the physiological activation of Ca2 + currents (Gurnett et al., 1996). Therefore, it is not surprising per se that GBP, as shown, indeed reduced Ca2 + currents. At present, very little evidence has been provided on this putative mechanism of action by GBP. We are aware of a single report (Wamil et al., 1991) describing the GBP-medi-

88

A. Stefani et al. / Neuropharmacology 37 (1998) 83–91

crease of the evokable Ca2 + current, does not shift the voltage-dependence of activation nor strongly changes the kinetic gating properties of the a1 channel, at least the A type (Gurnett et al., 1996). Moreover, in our hands GBP inhibited VGCC-channel through a fast, membrane-delimited mechanism which does not seem to involve an intracellular second messenger (consider the fastness of the response) or to interfere with membrane-bound phosphorilations which underlie facilitation (consider the negative experiments with double-pulse protocols, Dolphin, 1996). The GBP-induced inhibition of Ca2 + currents was studied in three cellular phenotypes, but the percentage block strongly differed in that it was much larger in the cortex than in the striatum and GP. This difference may be explained by different hypothesies. Firstly the presence, the distribution and the affinity of GBP binding sites may differ among the areas we analysed;

Fig. 4. Type of HVA Ca2 + channel involved in GBP-mediated modulation. (A – D) Representative test-activated Ca2 + currents (standard and facilitated currents) in control (A), under 10 mM GBP (B), under 5 mM nimodipine (C) and GBP plus nimodipine (D). Note that nimodipine, which almost completely prevents facilitation, occludes the GBP-induced response (D). (E) The histogram highlights the different prevalence of DHP-sensitive Ca2 + currents in the cortex, striatum and GP, as well as the negligible residual inhibition by GBP in the presence of nimodipine ( B 1%).

ated inhibition of the responses to the L-type Ca2 + channel agonist Bay K 8644 in mouse spinal cord (Wamil et al., 1991). Our findings are in clear agreement with that study. We have shown, however, that not only the L-type, but also the N and P/Q types, although to a much lesser degree, are affected by GBP. An important feature of the inhibition was that it did not modify strongly the activation and voltage-dependence of the conductance. Accordingly, Gurnett et al. (1996) have shown that the addition of the a2d subunit to mutant constructs, although producing a 9-fold in-

Fig. 5. Type of HVA Ca2 + channel involved in GBP-mediated modulation. (A) The partial block of the GBP response by v-CgTx GVIA in a pyramidal cortical neuron. (A) a, control; b, 30 mM GBP; c, 2 mM v-CgTx GVIA; b+c, 30 mM GBP in the presence of v-CgTx GVIA. (B) The histogram compares the average GBP-mediated inhibition of Ca2 + currents in control ( −33.5% 93.84, n= 4), in the presence of 2 mM v-CgTx ( − 15.1% of the residual current 92.24, n =4) and 2 mM v-CgTx MVIIC (−14.25% of the residual current 93.06, n = 4).

A. Stefani et al. / Neuropharmacology 37 (1998) 83–91

further biochemical studies are required to address these questions. Yet, however, the substantial similarity of the IC50 and minimal effective concentrations render this interpretation unlikely. The large difference observed in the responses might rely upon the peculiar sensitivity, elevated in one structure and negligible in others, of a channel subtype. Yet, this was not the case. As shown, DHP blocked the response in neurons isolated from the cortex as well as from the striatum or GP. Part of the discrepancy, however, may derive from the different prevalence of DHP-sensitive channels in the areas studied. In fact, L-type channels represent about 40% of the HVA Ca2 + currents in the neocortex (Stefani et al., 1996b), but less than 30% in the striatum and GP. Another hypothesis, at present speculative, involves the assembly and expression of functional Ca2 + channel complexes in situ in different areas (Gee et al., 1996). In other words, we should take into account the possibility that the channel structure itself, in cortical versus striatal or pallidal cells, is differently affected by the conformational changes promoted by the a2d subunit coassembly in physiological conditions and/or under pharmacological modulation. These issues need to be investigated by molecular biology approaches and by further pharmacological studies. Although small, the GBP-induced reduction of Ca2 + currents in the striatum and GP might be important regarding the putative impact of GBP on motor performances. The occurrence of involuntary movements, mainly dystonic, was recently described in patients treated with GBP (Buetefisch et al., 1996; Reeves et al., 1996). Conversely, the possible effectiveness of GBP in controlling restless legs syndrome (Mellick and Mellick, 1996) or postural and essential tremor (personal observations) is also under evaluation. These reports focus on the hypothesis that GBP may interact with endogenous dopamine, through a decrease of brain monoamine release, as already postulated in brain slices, or by a direct interference with the firing pattern of basal ganglia nuclei. The observed inhibition of Ca2 + currents by GBP is unlikely to be its unique mechanism of action, considering the rather conflicting reports on the utilization of DHP antagonists as anticonvulsants (Larkin et al., 1992). As previously mentioned, GBP has been suggested to modify GABA release and glutamate synthesis in 6i6o (Goldlust et al., 1995; Petroff et al., 1996). In addition, in a similar manner to several other AEDs, GBP promoted use-dependent inhibition of the sodium conductance, although this effect was noticeable only after a rather long-lasting incubation (\ 1 h) and at relatively high concentrations (\500 mM GBP) (Wamil and McLean, 1994). Significant aspects of the GBP-mediated decrease of Ca2 + currents we describe are the

89

fastness and reproducibility of the response as well as its efficacy in the low mM range. These observations suggest that GBP inhibition of Ca2 + conductances should play a central role in determining clinical responsiveness to GBP as well as to side effects. Conceivably, GBP might interact with the a2d subunit of the Ca2 + channel and with L-channels in general, not only in the central nervous system, but also in skeletal muscles and heart. Conversely, the GBP-induced inhibition of L- and non L-type VGCC suggests this anticonvulsant may be used not only for control of fast occurring Ca2 + -dependent hyperexcitability but also of slowly developing Ca2 + -dependent neurotoxicity. A similar role has also been suggested for the potent inhibition of L-type conductances by FBM, although this agent has not yet been utilized as an experimental neuroprotectant. On the contrary, GBP, whose clinical safety seems much larger, has already been tested in clinical trials for neurodegenerative diseases such as SLA (Gurney et al., 1996). Based upon our findings (occlusion by DHPs of the GBP response, quite partial antagonism by vCgTxGVIA and v-CgTxMVIIC), the Ca2 + channel subtypes which prevail in the axon terminal regions (N and P/Q channels) should not be modulated robustly by GBP. The precise impact of GBP on transmitter release, however, is not yet defined. In this regard, the pharmacological effects of GBP, in the presence or absence of agents which modulate VGCC, should be tested in preparations where the neuronal circuitry is preserved or epilepsy models are reproduced. The latter would help in elucidating the physiological effects that the GBP inhibition of Ca2 + current has on the cellular firing properties and in verifying whether GBP interacts with ligand-gated responses at the postsynaptic level.

Acknowledgements This work was supported by CNR Grants to AS, GB and by Ministero della Sanita` (‘Progetto Finalizzato’) to FS.

References Beydoun, A., Uthman, B.M., Sackellares, J.C., 1995. Gabapentin: pharmacokinetics, efficacy and safety. Clin. Neuropharmacol. 18, 469 – 481. Brust, P.F., Simerson, S., McCue, A.F., et al., 1993. Human neuronal voltage-dependent calcium channels: studies on subunit structure and role in channel assembly. Neuropharmacology 32, 1089– 1102. Buetefisch, C.M., Gutierrez, A., Gutman, L., 1996. Choreoathetotic movements: a possible side effect of gabapentin. Neurology 46, 851 – 852.

90

A. Stefani et al. / Neuropharmacology 37 (1998) 83–91

Calabresi, P., Stefani, A., DeMurtas, M., Pisani, A., Sancesario, G., Bernardi., G., 1995. Action of GP 47779, the active metabolite of oxcarbazepine, on the corticostriatal system. I. Modulation of corticostriatal synaptic transmission. Epilepsia 336, 990–996. Calabresi, A., Siniscalchi, A., Pisani, A., Stefani, A., Mercuri, N.B., Bernardi., G., 1996. A field potential analysis on the effects of lamotrigine, GP 47779 and felbamate in neocortical slices. Neurology 47, 557 – 562. Catterall, W.A., 1995. Structure and function of voltage-gated ion channels. Annu. Rev. Biochem. 64, 493–531. Choi, S., Lovinger, D.M., 1996. Metabotropic glutamate receptor modulation of voltage-gated Ca2 + channels involves multiple receptor subtypes in cortical neurons. J. Neurosci. 16, 38– 45. Cochran, J.W., 1996. Restless legs syndrome. J. Am. Med. Assoc. 275, 187. Dolphin, A.D., 1996. Facilitation of currents in excitable cells. Trends Neurosci. 19, 35 – 43. Ellis, S.B., Williams, M.E., Ways, N.R., et al., 1988. Sequence and expression of mRNAs encoding the alfa 1 and alfa 2 subunit of a DHP-sensitive calcium channel. Science 241, 1661–1664. Gee, N.S., Brown, J.P., Dissanayake, V.U.K., Offord, J., Thurlow, R., Woodruff, G.N., 1996. The novel anticonvulsant drug, Gabapentin (Neurontin), binds to the a2d subunit of a calcium channel. J. Biol. Chem. 271, 5768–5776. Goldlust, A., Su, T.Z., Welty, D.F., Taylor, C.P., Oxender, D.L., 1995. Effects of anticonvulsant drug gabapentin on the enzymes in metabolic pathways of glutamate and GABA. Epilepsy Res. 22, 1–11. Gurnett, A.C., De Waard, M., Campbell, P.K., 1996. Dual function of the voltage-dependent Ca2 + channel a2d subunit in current stimulation and subunit interaction. Neuron 16, 431–440. Gurney, M.E., Cuttings, F.B., Zhai, P., Doble, A., Taylor, C.P., Andrus, P.K., Hall, E.D., 1996. Benefit of vitamin E, Riluzole and Gabapentin in a transgenic model of familail amyotrophic lateral sclerosis. Ann. Neurol. 39, 147–157. Honmou, O., Kocsis, J.D., Richerson, G.B., 1995. Gabapentin potentiates the conductance incraese induced by nipecotic acid in CA1 pyramidal neurns in vitro. Epilepsy Res. 20, 193–202. Isom, L.L., De Jongh, K.S., Catterall, W.A., 1994. Auxiliary subunits of voltage-gated ion channels. Neuron 12, 1183–1194. Kanthasamy, A.G., Vu, T.Q., Yun, R.J., Truong, D.D., 1996. Antimyoclonic effect of gabapentin in a posthypoxic model of myoclonus. Eur. J. Pharmacol. 297, 219–224. Kocsis, J.D., Honmou, O., 1994. Gabapentin increases GABA-induced depolarization in rat neonatal optic nerve. Neurosci. Lett. 169, 181 – 184. Larkin, J.G., Besag, F.M.C., Cox, A., Brodie, M.J., 1992. Nifedipine for epilepsy? A double-blind, placebo-controlled study. Epilepsia 33, 346 – 352. Leiderman, D.B., 1994. Gabapentin as add-on therapy for refractory partial epilepsy: results of five placebo-controlled trials. Epilepsia 35, S74 – S76. Lorenzon, N.M., Foehring, R.C., 1995. Characterization of pharmacologically identified voltage-gated calcium channels in acutely isolated rat neocortical neurons. I Adult neurons. J. Neurophysiol. 73, 1430 – 1442. Macdonald, R.L., Kelly, M.K., 1994. Mechanisms of action of currently prescribed and newly developed antiepileptic drugs. Epilepsia 35, S41 – S50. Macdonald, R.L., Greenfield, J.L. Jr., 1997. Mechanisms of action of new antiepileptic drugs. Curr. Opin. Neurol. 10, 121–128. McLean, M.J., Macdonald, R.L., 1983. Multiple actions of phenytoin on mouse spinal cord neurons in cell culture. J. Pharm. Exp. Ther. 227, 779 – 789.

Meldrum, B.S., 1996. Update on the mechanism of action of antiepileptic drugs. Epilepsia 37, S4 – S11. Mellick, G.A., Mellick, LB., 1996. Management of restless legs syndrome with gabapentin (Neurontin). Sleep 19, 224 – 226. Mikami, A., Imoto, K., Tanabe, T., et al., 1989. Primary structure and functional expression of the cardiac dihydropyridine-sensitive calcium channel. Nature 340, 230 – 233. Patel, J., Naritoku, D.K., 1996. Gabapentin for the treatment of hemifacial spasms. Clin. Neuropharmacol. 19, 185 – 188. Petroff, O.A., Rothman, D.L., Behar, K.L., Lamoureux, D., Mattson, R.H., 1996. The effect of gabapentin on brain gammaaminobutyric acid in patients with epilepsy. Ann. Neurol. 39, 95 – 99. Pisani, A., Stefani, A., Siniscalchi, A., Mercuri, N.B., Bernardi, G., Calabresi, P., 1996. Electrophysiological actions of felbamate on rat striatal neurons. Br. J. Pharmacol. 116, 2053 – 2061. Reeves, A.L., So, E.S., Sharbrough, F.W., Krahn, L.E., 1996. Movement disorders associated with the use of gabapentin. Epilepsia 37, 988 – 990. Sayer, R.J., Brown, A.M., Schwindt, P.C., Crill, W.E., 1993. Calcium currents in acutely isolated human neocortical neurons. J. Neurophysiol. 69, 1596 – 1606. Segal, A.Z., Rordorf, G., 1996. Gabapentin as a novel treatment for postherpetic neuralgia. Neurology, 1175 – 1176. Siesjo, B.K., Bengtsson, F., 1989. Calcium fluxes, calcium antagonists and calcium-related pathology in brain ischemia, hypoglycemia and spreading depression: a unyfying hypothesis. J. Cereb. Blood Flow Metab. 9, 127 – 140. Siniscalchi, A., Bonci, A., Mercuri, N.B., Bernardi, G., 1996. Effect of riluzole on rat cortical neurons: an in vitro electrophysiological study. Br. J. Pharmacol. (in press). Stefani, A., Pisani, A., Mercuri, N.B., Bernardi, G., Calabresi, P., 1994. Activation of metabotropic glutamate receptors inhibits calcium currents and GABA-mediated synaptic potentials in striatal neurons. J. Neurosci. 14, 6734 – 6743. Stefani, A., Pisani, A., De Murtas, M., Mercuri, N.B., Marciani, M.G., Calabresi, P., 1995. Action of GP 47779, the active metabolite of oxcarbazepine, on the corticostriatal system. II Modulation of high-voltage-activated calcium currents. Epilepsia 336, 997 – 1002. Stefani, A., Spadoni, F., Bernardi, G., 1996a. Lamotrigine inhibits calcium currents in cortical neurons: functional implications. Eur. J. Pharmacol. 307, 113 – 116. Stefani, A., Calabresi, P., Pisani, A., Mercuri, N.B., Siniscalchi, A., Bernardi, G., 1996b. Felbamate inhibits dihydropyridine-sensitive calcium channels in central neurons. J. Pharm. Exp. Ther. 277, 121 – 129. Stefani, A., Spadoni, F., Bernardi, G., 1997a. Voltage-activated calcium channels: tartget of antiepileptic drug therapy? Epilepsia 38, 959 – 965. Stefani, A., Spadoni, F., Bernardi, G., 1997b. Differential inhibition of inward currents by lamotrigine, riluzole and phenyotin in cortical neurons: implications for neuroprotective strategies. Exp. Neurol. 747, 115 – 122. Stefani, A., Spadoni, F., Bernardi, G., 1997c. Group III mGluR agonists modulate HVA Ca2 + currents in pyramidal neurons of the adult rat. Exp. Brain Res. (in press). Surmeier, D.J., Seno, N., Kitai, S.T., 1994. Acutely isolated neurons of the rat globus pallidus exhibit four types of high-voltage-activated Ca2 + currents. J. Neurophysiol. 71, 1272 – 1280. Takahashi, T., Momiyama, A., 1993. Different types of calcium channels mediate central synaptc transmission. Nature 366, 156– 158. Taylor, C.P., Vartanian, M.G., Yuen, P.W., Bigge, C., SumanChauan, N., Hill, D.R., 1993. Epilepsy Res. 14, 11 – 15.

A. Stefani et al. / Neuropharmacology 37 (1998) 83–91 Tymianski, M., tator, CH, 1996. Normal and abnormal calcium homeostasis in neurons: a basis for the pathophysiology of traumatic and ischemic central nervous system injury. Neurosurgery 38, 1176 – 1195. Wamil, A.W., McLean, M.J., Nashville, T.N., Taylor, C.P., 1991. Neurology 41 (S1), 140. Wamil, A.W., McLean, M.J., 1994. Limitation by gabapentin of high frequency action potential firing by mouse central neurons in cell culture. Epilepsy Res. 17, 1–11.

.

91

Welty, D.F., Selielke, G.P., Rothstein, J.D., 1995. Potential treatment of amyotrophic lateral sclerosis with gabapentin: a hypothesis. Ann. Pharmacother. 29, 1164 – 1167. Williams, M.E., Brust, P.F., Feldman, D.H., et al., 1992. Structure and fucntional expression of an v-conotoxin-sensitive human N-type calcium channel. Science 257, 389 – 395. Witcher, D.R., De Waard, M., Sakamoto, J., Franzini-Armstrong, C., Pragnell, M., Kahl, S.D., Campbell, K.P., 1993. Subunit identification and reconstitution of the N-type Ca2 + channel complex purified from brain. Science 261, 486 – 489.