Different Effects of Gabaergic Anticonvulsants on 4-Aminopyridine-Induced Spontaneous Gabaergic Hyperpolarizations of Hippocampal Pyramidal Cells—Implication for their Potency in Migraine Therapy

Abstract

Clinical studies indicate anti-migraneous efficacy of the probably GABAergic anti-convulsants valproate and gabapentin. For the GABAergic anticonvulsants vigabatrin and tiagabine, studies about antimigraneous efficacy are missing. The aim of this study was to test the GABAergic potency of these drugs in vitro before further clinical studies. Intracellular recordings were obtained from hippocampal pyramidal cells. Spontaneous GABAergic hyperpolarizations (SGH) elicited by 75 μ m 4-aminopyridine were used to test the effect of these drugs on GABA-dependent potentials. Tiagabine (0.1 m m) prolonged the duration of SGH. Furthermore, monophasic SGH turned over into triphasic typical GABAergic membrane potential fluctuations within 20 min. In contrast, valproate, gabapentin, and vigabatrin failed to affect SGH up to 60 min of application. The reason for the fast action of tiagabine on SGH may be caused by a faster increase of synaptic GABA levels compared with other drugs. As migraine therapy benefits from an augmentation of GABA activity, we recommend clinical studies of tiagabine as a fast-acting agent in migraine attacks.

Introduction

Enhancing the cerebral GABA system is one of many proposed mechanisms to limit convulsive seizures. Furthermore, enhanced GABA potentials are also discussed as one possible mechanism of valproate's action in migraine prophylaxis (1). While the anticonvulsant potency of the probably GABAergic substances valproate and gabapentin (2) are proven, first clinical trials indicate that migraine patients benefit from the prophylactic treatment with either drug, too (3–8). The GABAergic anticonvulsants vigabatrin and tiagabine have not yet been studied with respect to their anti-migraneous effectiveness (9). Table 1 summarizes commonly postulated mechanisms of how GABAergic anti-convulsants can act in epileptic tissue. Taken together, in addition to the GABAergic mechanisms, valproate as well as gabapentin act on various ion channels and transmitter systems, while for vigabatrin and tiagabine only GABAergic mechanisms are described. Both drugs increase GABA levels in brain but interact with different cellular components: vigabatrin by irreversible inhibition of the GABA-degrading enzyme GABA aminotransferase (26), and tiagabine by reversible inhibition of a GABA re-uptake via high-affinity binding to the GABA carrier, GAT-1 (27).

Table 1

Postulated mechanisms of tested anti-convulsants

	GABAergic mechanisms	Other mechanisms
Valproate	Activation of GABA-synthesizing enzyme glutamic acid decarboxylase (10) Inhibition of GABA-degrading enzymes GABA aminotransferase (10, 11), succinate semialdehyde dehydrogenase (12)	Inhibition of release of excitatory amino acid aspartate (13) Increase of extracellular concentration of hydroxybutyrate, serotonin, dopamine, glycine, taurin, homovanillic acid (14–19) Depression of inward sodium and calcium currents (20, 21) Increase of potassium conductance (22)
Gabapentin	Activation of GABA-synthesizing enzyme glutamic acid decarboxylase (23) Inhibition of GABA-degrading enzyme GABA aminotransferase (24)	Activation of glutamate-degrading enzyme glutamate dehydrogenase (23) Inhibition of glutamate-synthesizing enzyme amino acid aminotransferase (23, 24) Binding to α2δ-subunit of l-type calcium channel (25)
Vigabatrin	Irreversible inhibition of GABA-degrading enzyme GABA aminotransferase (26)
Tiagabine	Reversible inhibition of GABA uptake by binding to the GABA uptake carrier, GAT-1 (27)

In the present study we used GABAergic hyperpolarizations experimentally induced by 4-aminopyridine(4-AP) which is commonly used to elicit hyperexcitation in cortical neurones (28). 4-AP evokes spontaneous GABAergic events in hippocampal pyramidal neurones, as previously shown by several investigators (29–32). This was interpreted to result from a 4-AP-induced inhibition of special potassium channels of GABAergic interneurones whose pronounced firing rates are likely to induce spontaneous GABAergic events in post-synaptic pyramidal neurones.

Compared with a local application of GABA (33), the 4-AP-treated hippocampal slice is a very convenient model to study the potency of GABAergic drugs whose different modes of action on the GABA system are of special interest concerning the pathophysiology and the treatment of migraine.

Methods

Transverse hippocampal slices (200–400 μm thick) were prepared from brains of ether-anaesthetized adult guinea pigs (300–400 g). Slices were preincubated for 2 h in 28°C warm saline equilibrated with 5% CO₂ in O₂. This saline contained (in mm): NaCl 124, KCL 3, CaCl₂ 0.75, MgSO₄ 1.3, KH₂ PO₄ 1.25, NaHCO₃ 26, and glucose 10. After preincubation, slices were transferred to a perspex recording chamber (volume 4 ml) which was mounted on an inverted microscope. In this chamber the submerged slices were continuously superfused at a rate of 4 ml/min by 32°C warm saline. The composition of this control solution (CTR) was the same as during the preincubation period except for the calcium concentration which was raised to 1.75 mm.

The following drugs were added to CTR: 4-AP (75 μm; Sigma, Deisenhofen, Germany), DL-2-amino-5-phosphonovalerate (APV, 50 μm; Sigma), 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, 50 μm; Sigma), bicuculline-methiodide (BMI, 50 μm; Sigma), 2-hydroxysaclofen (HS, 300 μm; Sigma), valproate (1–2 mm; Desitin, Hamburg, Germany) gabapentin (1–2 mm; Parke Davis, Berlin, Germany), vigabatrin (0.1–0.8 mm; Hoechst, Bad Soden, Germany), and tiagabine (0.1 mm; Sanofi Winthrop, München, Germany).

Intracellular recordings were obtained from the somata of CA3-neurones with sharp glass microelectrodes filled with 2 m potassium methylsulphate.

Results

In this study we investigated the effect of the GABAergic anticonvulsants on spontaneous GABAergic hyperpolarizations (SGH). Therefore, intracellular recordings were obtained from 47 CA3-neurones of 45 slices. Only CA3-neurones were selected which had AP-amplitudes > 50 mV and a resting membrane potential <−50 mV. In all tested 47 CA3-neurones epileptiform activity consisted of periodically appearing bursts whichdeveloped during application of 75 μm 4-AP.

SGH induced by 4-AP

SGH occurred in 28 out of 47 CA3-neurones. These SGH had a duration of 2–4 s. The hyperpolarizing changes in membrane potential were strictly monophasic and reversal potentials of −58 to −67 mV were measured. The SGH persisted during the blockade of NMDA- and AMPA-receptors with 50 μm APV and 50 μm CNQX, but could be reduced with 50 μm BMI and 300 μm HS (data shown in (32)). Thus, these hyperpolarizations represent post-synaptic endogenous GABAergic hyperpolarizations which were probably due to an augmented release of GABA from GABAergic interneurones induced by 4-AP (31, 34).

Effects of valproate, gabapentin, vigabatrin on SGH

In a first step those drugs acting on GABA-degrading or-synthesizing enzymes were tested. The testing period was limited to 60 min, which is a reasonable duration for intracellular recordings of CA3-neurones.

In seven neurones SGH remained unchanged in amplitude and duration when slices were exposed to 1–2 mm valproate. SGH also persisted during exposure of 1–2 mm gabapentin (n = 5) and 0.1–0.8 mm vigabatrin (n = 6).

Effects of tiagabine on SGH

In contrast to the above mentioned drugs the GABA re-uptake inhibitor tiagabine clearly changed SGH: within 10–15 min after washin, the duration of SGH increased by 50–70%. Thereafter, the monophasic SGH converted into triphasic membrane potential fluctuations with an early hyperpolarization (component a), a following depolarization (component b), and a late hyperpolarization (component c) in all neurones (n = 10, Fig. 1a). This triphasic event was completely reversible within 30 min after the onset of washout.

Figure 1

(a) Conversion of monophasic spontaneous GABAergic hyperpolarizations (SGH) induced by 4-aminopyridine (4-AP) into triphasic membrane potential fluctuations (a,b,c) under tiagabine 0.1 mm. (b) Development of triphasic membrane potential fluctuations induced by ejection of increasing amounts of GABA from a glass pipette to the soma of CA3-neurones (taken in changed form from (33)).

Discussion

The objective of the study was to compare the GABAergic potency of the four anticonvulsants valproate, gabapentin, vigabatrin, and tiagabine. As a parameter for assessment of GABAergic potency, SGH induced by 4-AP were selected. In contrast to valproate, gabapentin and vigabatrin, only tiagabine prolonged the duration of the SGH after 10–15 min of application.

This observation was supported by the study of Thompson & Gähwiler (35), who also described a clearly prolonged duration of evoked mono- and polysynaptic inhibitory post-synaptic potentials (IPSP) of CA3-neurones exposed to tiagabine (10–25 μm). They postulated that part of the prolonged duration of IPSP result from a GABA_B receptor-mediated component. Whole cell recordings revealed that tiagabine also augmented GABA_A receptor-mediated responses (36, 37).

When 0.1 mm tiagabine was given to the superfusate, the monophasic SGH converted into triphasic deflections of the membrane potential. In contrast, valproate, gabapentin and vigabatrin failed to change SGH. The observed conversion from monophasic to triphasic potentials during tiagabine corresponds to the responses obtained during external application of increasing GABA amounts by iontophoresis or pressure pulses to the soma of hippocampal pyramidal cells ((33, 38, 39) and compare Fig. 1b). Therefore, we strongly suggest that the effect seen with tiagabine was due to a growing level of GABA in and around synaptic clefts.

There are several lines of evidence that GABergic drugs modulate biochemical and physiological events involved in the pathophysiology of migraine. Spreading depression, which is a cortical event that underlies the migraine aura, may be suppressed by an increase of inhibitory GABAergic neurotransmission induced, for example, by valproate (1). The release of vasoactive neuropeptides (substance P, neurokinin A, calcitonin-gene-related peptide) from activated sensory nerve terminals (40) can be decreased by valproate (41). Furthermore, the activation of the trigeminal nucleus caudalis, responsible for modulation of the nociceptive signals, can be suppressed by GABAergic drugs. Expressing of immediate early gene c-fos as a marker of neuronal activity within trigeminal nucleus caudalis was reduced by valproate (42) as well as allopregnanolone, a progesterone metabolite which modulates GABA_A receptor activity (43). The effect of valproate could be reversed by bicuculline (42). Together, these studies suggest that an augmentation of the inhibitory central GABA system plays a pivotal role in suppressing migraine-related events.

We conclude that the re-uptake inhibitor tiagabine augments the level of synaptic GABA faster compared with the other drugs which work on enzymatic pathways. Keeping in mind tiagabine, administered intraperitoneally, increases extracellular GABA overflow in the globus pallidus of Sprague-Dawley rats with peak values at about 40 min after application (44), this drug appears well suited as a fast-acting anti-migraine drug. The clinical efficacy of tiagabine in migraine therapy, especially in acute migraine attacks, should be investigated.