Focal and Secondarily Generalised Convulsive Status Epilepticus Induced by Thiocolchicoside in the Rat

Seizure 2003; 12: 508–515 doi:10.1016/S1059–1311(03)00053-0

Focal and secondarily generalised convulsive status epilepticus induced by thiocolchicoside in the rat

GIANPIETRO SECHI ‡, PIERLUIGI DE RIU †, OMBRETTA MAMELI §, GIOVANNI A. DEIANA ‡, GIOVANNI A. COCCO ‡ & GIULIO ROSATI ‡

†Department of Traumatology, Orthopedy and Occupational Health, University of Torino, Torino, Italy; ‡Neurological Clinic, Viale S. Pietro, 10, Sassari, Italy; §Department of Biomedical Sciences, University of Sassari, Sassari, Italy

Correspondence to: GianPietro Sechi, M.D., Neurological Clinic, Viale S. Pietro, 10, 07100 Sassari, Italy. E-mail: [email protected]

The objective of this study was to document the convulsant properties of thiocolchicoside in rats, and to characterise the electroclinical pattern of epileptic seizures. Experiments were carried out in three groups of male Wistar rats: in group A, thiocolchicoside was applied topically to the pia, or given by microinjection to the cerebral cortex (2 µg/µl); in group B, the drug was administered parenterally (6 mg/kg) to rats with minimal lesions of the dura and arachnoid membranes; in group C, thiocolchicoside was administered parenterally (up to 12 mg/kg) to intact rats. In all animals, electroclinical activity was continuously monitored for at least 3 hours after thiocolchicoside injection or application. In group A, electrographic and behavioural activity of focal motor seizures occurred in 100% of animals, developing into a focal status epilepticus; in group B, a multifocal epileptic pattern with secondary generalisation, clinically characterised by clonic or tonic–clonic seizures occurred in 100% of animals, until a secondarily generalised convulsive status epilepticus; in group C, none of animals showed either electrographic or behavioural seizure activity. Our study documents that thiocolchicoside has a powerful convulsant activity in the rat, perhaps due to an antagonistic interaction of the compound with a cortical subtype of the GABAA receptor. © 2003 BEA Trading Ltd. Published by Elsevier Science Ltd. All rights reserved.

Key words: thiocolchicoside; status epilepticus; GABAA receptors.

INTRODUCTION side. In order to induce focal status epilepticus, this drug was applied topically to the pia or given by Thiocolchicoside is a semi-synthetic derivative of a microinjection to the cerebral cortex. To induce a naturally occurring colchicoside, a colchicine ana- secondarily generalised status epilepticus, it was ad- logue, extensively used clinically in some countries ministered parenterally to rats with minimal lesions as central acting muscle-relaxant, anti-inﬂammatory of the dura and arachnoid membranes. and analgesic (Fig. 1)1. Colchicine may induce experimental epileptic foci in animals (mice, rabbits and rats)2, and its intracranial administration causes METHOD generalised convulsions and death3. Our preliminary experimental ﬁndings, presented previously as an Studies were performed under a protocol approved abstract4, and our recent clinical data5 indicate that by the Institutional Animal Care Committee. Male thiocolchicoside also has powerful convulsant activity rats, Wistar Morini strain, weighing 270–350 g (3–4 in animals and in humans; particularly, in neurological months of age), fasting for 12 hours before the experi- patients with disruption of their blood–brain barrier ments, and kept at a room temperature for 22 ◦C, were (BBB) and in patients with history of epilepsy5.In anaesthetised with ether or ketamine hydrochloride this paper, we document the occurrence of focal and (about 50 mg/kg, intraperitoneally), the head mounted secondarily generalised convulsive status epilepticus in a stereotactic apparatus. Animals were randomly in the rat following the administration of thiocochico- divided into three groups (groups A, B and C).

1059–1311/$30.00 © 2003 BEA Trading Ltd. Published by Elsevier Science Ltd. All rights reserved. Convulsant activity of thiocolchicoside 509 Fig. 1: Structures of thiocolchicoside (I), colchicoside (II) and colchicine (III). 510 G. Sechi et al.

The rats of group A were further divided into two (BLA) (referred to bregma: 2.8 mm posterior, 8.5 mm subgroups: in the first, after ketamine anaesthesia, the ventral, 5.0 mm lateral to the midline) (five animals) cerebral cortex was exposed by craniotomy, the dura and gigantocellular proper mesencephalic reticular was removed and electrocorticographic (ECoG) pat- formation (MRF-Gi) (referred to bregma: 10.8 mm terns recorded bipolarly by six silver-ball electrodes posterior, 9.5 mm lateral to the midline, 10 mm ven- applied to the pia in bilateral symmetrical position (the tral) (five animals). reference electrode was a screw placed on the nasal In group C, thiocolchicoside dissolved in saline bone). Focal epileptic seizures were induced in the was administered i.p. or i.m. to intact rats, at doses exposed cortex (right or left) by application of a small of 6 mg/kg (five animals), 8 mg/kg (five animals) piece of Spongostan (2 mm × 2 mm), a sponge of fib- and 12 mg/kg (five animals). The surface EEG was rin soaked in a 2 mg/ml isotonic sodium chloride so- recorded by means of six subcutaneous needle elec- lution of thiocolchicoside (15 animals). In the second trodes inserted in bilateral symmetrical position (the subgroup, focal seizures were induced by one intra- reference electrode was fixed on the pinna). cortical microinjection of thiocolchicoside dissolved In all animals the exposed cortex was covered with in saline (the injection volume used was 1 µl and the warmed paraffin oil at 37 ◦C. Body temperature was total amount of thiocolchicoside applied was 2 µg), maintained between 36.5 and 37.5 ◦C with a heating near the recording electrodes (10 animals). In this lat- pad. Both depth and surface electrodes were connected ter subgroup, each animal underwent a tracheotomy to an EEG recording apparatus (OTE Biomedica). The under ether anaesthesia, and was then immobilised electrocardiogram was routinely monitored. All ani- with gallamine and artificially ventilated. Novocaine mals were continuously monitored for at least 3 hours infiltration of all warmed edges and pressure points after the thiocolchicoside injection or application. At was carefully carried out to minimise possible source least 10 minutes of basal ECoG or EEG recording was of pain. made in each animal. The position of the depth elec- In group B, after ketamine anaesthesia, thiocolchi- trodes was checked. Data given are mean ± standard coside dissolved in saline was administered either deviation. into the peritoneum (i.p.), or intramuscularly (i.m.) (6 mg/kg) to rats with minimal lesions of the dura and arachnoid membranes. The lesions of these mem- RESULTS branes occurred during the insertion of the screw electrodes into the skull, or during craniotomy. In In group A, treated with focal cortical thiocolchico- particular, the lesions made during the insertion of the side, electrographic and behavioural seizure activity screw electrodes were four, in symmetrical positions occurred in 100% of animals. The latency to focal in the area of the frontoparietal cortex bilaterally. spike activity was 3 ± 2 minutes following topical ap- The extent of the lesions induced by screw elec- plication of the drug on the pial surface, and 2.0 ± trodes or craniotomy was about 0.1mm × 0.1mm 1 minute following one intracortical microinjection. and 2 mm × 2 mm, respectively. In preliminary exper- The frequency of spikes was 48 ± 20 minutes. Af- iments, different amounts of thiocolchicoside were ter about 60 minutes, spikes and polyspikes, intermin- injected i.p. or i.m. in rats of the same age and weight gled with both bursts of focal recruiting discharges of in order to establish the necessary dose of the drug 1.5- to 3-second duration and intermittent ictal focal needed to induce constant and reliable electroclinical seizures of several seconds duration occurred, accom- responses. Similarly, the extent of the lesions of dura panied by controlateral myoclonic jerks (Fig. 2). The and arachnoid membranes needed to obtain constant electrographic seizure activity was interrupted by a and reliable seizure activity was established in prelim- near complete depression of cortical activity on ECoG. inary experiments. The surface electroencephalogram This activity progressed into a focal status epilepticus (EEG) was recorded from the cortical convexity ei- and lasted for at least 3 hours. ther by means of six screw electrodes inserted into In group B, treated with parenteral thiocolchico- the skull in bilateral symmetrical positions (the refer- side, a multifocal epileptic pattern with secondary ence electrode placed on the nasal bone) (15 animals) generalisation, clinically characterised by clonic or or, when a wide craniotomy was performed, by using tonic–clonic seizures occurred in 100% of animals. six ball electrodes applied to the pial surface (the The latency to first spike activity and to the first sec- reference electrode was a screw electrode placed on ondarily generalised seizure until a convulsive status the nasal bone) (10 animals). In some cases bipolar epilepticus developed was 3 ± 1 and 60 ± 10 minutes, concentric electrodes were stereotaxically implanted respectively. After thiocolchicoside injection, the into subcortical structures according to the atlas of spikes rose from the opposite hemisphere in a com- Paxinos and Watson6. The following structures were pletely asynchronous way for 2–4 minutes, since when explored: left basolateral amygdaloid nucleus anterior a progressive and rapid temporal alliance between Convulsant activity of thiocolchicoside 511

Fig. 2: Development of focal epileptiform activity in the left rat neocortex, after topical application of a small piece of Spongostan soaked in an isotonic solution of thiocolchicoside. (A)–(E ) respectively: basal recordings, 0.5, 1.5, 10 and 60 minutes after drug application; L, left. Vertical calibration 200 µV; horizontal calibration 1 second. 512 G. Sechi et al.

Fig. 3: Development of cortical epileptiform activity, until a secondary generalised seizure in a rat, with minimal lesions of dura and arachnoid membranes, injected with 6 mg/kg thiocolchicoside. (A)–(E) respectively: basal recordings, 3, 5, 30 and 60 minutes after drug injection; L, left. Vertical calibration 200 µV; horizontal calibration 1 second. the epileptic activity of the two hemispheres was clonic or, occasionally tonic–clonic, which lasted observed, giving origin to a pseudogeneralised pat- 20–180 seconds (Fig. 3). The interval of intermittent tern. The amplitude and frequency of spikes rapidly generalised seizure activity was 60 ±20 minutes. This increasing, up to 0.6 mV and 50/minute, respectively. activity lasted for at least 3–5 hours, and resulted in After about 1 hour, the epileptic potentials became death in about 5% of treated rats. Depth electrode subcontinuous and characterised by spike trains un- recordings initially showed no involvement of the ex- til ictal electroclinical generalised seizures mainly plored structures. With an 8- to 15-minute delay, as to Convulsant activity of thiocolchicoside 513

Fig. 4: Development of cortical epileptiform activity in a rat, with minimal lesions of dura and arachnoid membranes, injected with 6 mg/kg thiocolchicoside. (A)–(D) respectively: basal recordings, 3, 5 and 20 minutes after drug injection. The paroxysmal activity appears in MRF-Gi with 15-minute delay, as to cortex. MRF recordings are given in addition to cortical recordings. L, left. Vertical calibration 200 µV; horizontal calibration 1 second. the cortex, paroxysmal activity appeared also in sub- DISCUSSION cortical structures, at first in BLA, then in MRF-Gi. At this time, a high degree of synchrony of spiking Thiocolchicoside, colchicoside and colchicine share between the various structures was evident (Fig. 4). the same benzo[α]heptalenic moiety, however, thio- In group C, at the doses tested, none of animals colchicoside differs from colchicoside and colchicine showed either electrographic or behavioural seizure by substitution of a methoxy group for a thiomethyl activity. group on ring A (Fig. 1)1. Structural analysis of 514 G. Sechi et al. these compounds reveals the need for at least a single and by the finding that neurosteroids modulate the methoxy group on rings A and C for neurotoxicity binding in cortical but not in spinal cord–brainstem of colchicine2. This drug, indeed, binds to tubulin synaptic membranes9. In particular, at cortical level, and has been found to be a selective toxin for den- thiocolchicoside seems to recognise a subpopulation 2 tate granule cells in the hippocampus of rats . The of GABAA receptors, expressing low-affinity binding concomitance of a methoxy group on rings A and sites for GABA9. Since the low-affinity recognition C, instead, does not seem necessary for the convul- site seems to be an antagonistic-preferring site7, this sant activity of colchicine2. Our data on the powerful finding may readily explain both the electrophysio- convulsant activity of thiocolchicoside (lacking a logical results reported in this study and the powerful methoxy group on ring A) agree with these findings. convulsant activity of thiocolchicoside noticed in Our study documents that thiocolchicoside given at humans5. doses of 6–12 mg/kg is unable to induce, in intact Another interesting finding concerns the potency of rats, any electrical or behavioural paroxysmal activity. strychnine in inhibiting thiocolchicoside binding, par- However, when this compound is applied topically to ticularly in the cerebral cortex9, suggesting a probable the pia, given by microinjection to cerebral cortex, or interaction of both compounds at the same binding administered parenterally at doses of 6 mg/kg to rats sites9 at cortical level. Strychnine, it has recently been with minimal lesions of the dura and arachnoid mem- shown, may interact in the brain, at submicromolar branes, it displays within a few minutes, a powerful doses, with strychnine-sensitive glycine receptors, convulsant activity. GABAA receptors, as well as nicotinic acetylcholine The impermeability of the BBB to thiocolchicoside receptors8–10, which are homologous members of a is likely related to the dose used since in preclinical receptor superfamily11, 12. studies1 it was noted that in mice, at doses of 20 mg/kg Like strychnine, thiocolchicoside has been shown to of thiocolchicoside, some animals exhibited trembling interact with GABAA receptors and strychnine-sensi- and died 40 minutes after dosing. The relative imper- tive glycine receptors8, 9. These latter receptors are meability of the BBB to therapeutic doses of thio- mainly represented in the spinal cord and brainstem13, colchicoside may explain why the convulsant activity and might be involved in the myorelaxant activity 8 of this drug in humans so far has not been recognised, of thiocolchicoside . GABAA receptors, instead, are despite its extensive use in clinical practice for about mainly represented in the cerebral cortex14, 15, and 30 years5. may be involved both in the convulsant activity of thio- In our experiments, after topical microinjection, colchicoside, and in spiking produced by strychnine thiocolchicoside was able to induce focal cortical in the cortex16. In particular, the convulsant activity seizures in micromolar amounts and after a short la- of thiocolchicoside is likely to involve an overcom- tent period, while after its parenteral administration, ing effect of thiocolchicoside antagonism at a subset cortical spikes occurred firstly before subcortical of GABAA receptors, on thiocolchicoside inhibitory structures. These findings, in our opinion, strongly effects of allosteric activation of strychnine-sensitive point to the forebrain, and more specifically the cor- glycine receptors. The prevalence of GABAA recep- tex, as the likely site where thiocolchicoside-induced tors in the cerebral cortex, with respect to strychnine- seizures are initiated. The occurrence of the high- sensitive glycine receptors, fits this possibility13–15. est levels of specific thiocolchicoside binding in the According to Engel17, in epilepsy research the thio- deep layers of the frontal cortex of the rat fits this colchicoside model would be classified with the group possibility7. of ‘acute experimentally induced seizure models’, The mechanism of action of thiocolchicoside is which would be equivalent to reactive seizures in hu- only partially understood. Previous studies suggested mans, or with the group of animal models of epileptic that an agonistic interaction of thiocolchicoside with seizures in nonepileptic animals reported by Löscher spinal-strychnine-sensitive glycine receptors could in a recent review18. The electroclinical pattern was mediate its myorelaxant activity8; such interaction, consistent with an acute focal and secondarily gen- however, does not readily explain how it may induce eralised convulsive status epilepticus. In this model, seizures5. the occurrence of a minimally disrupted BBB par- Recent studies have, instead, demonstrated the allels a condition that can be found frequently in presence of thiocolchicoside-specific binding sites in clinical practice (e.g. after head trauma), in which the rat cerebral cortex and in spinal cord–brainstem9, it has been shown that thiocolchicoside may induce with a pharmacological profile indicating a prefer- epileptic seizures in humans, at very low doses5.In ential interaction of this compound with a cortical comparison with similar models characterised in far 9 subtype of the GABAA receptor . This peculiar in- greater detail (e.g. those obtainable with bicuculline teraction is suggested both by the different affinities or strychnine or kainate)17, 18, the thiocolchicoside for the GABAA receptor agonists and antagonists, model shows the same reproducibility and consistency Convulsant activity of thiocolchicoside 515 and a peculiar mechanism of action (i.e. an antago- 5. De Riu, P. L., Rosati, G., Sotgiu, S. and Sechi, G. P. Epileptic seizures after treatment with thiocolchicoside. Epilepsia 2001; nistic interaction with a subset of GABAA receptors, 42: 1084–1086. and an agonistic interaction with strychnine-sensitive 6. Paxinos, G. and Watson, C. The Rat Brain in Stereotaxic 7, 9 glycine receptors) . This model could have utility Coordinates, 2nd edn. San Diego, Academic Press, 1987. in studies of the functional anatomy, neural networks 7. Balduini, W., De Angelis, V., Mazzoni, E., Depoortere, H., and neurochemical mechanisms involved in epilepsy Cattabeni, F. and Cimino, M. Autoradiographic localization of [3H]thiocolchicoside binding sites in the rat brain and spinal and epileptic seizures. cord. Neuropharmacology 2001; 40: 1044–1049. 8. Cimino, M., Marini, P. and Cattabeni, F. Interaction of thiocolchicoside with [3H]strychnine binding sites in rat spinal CONCLUSIONS cord and brainstem. European Journal of Pharmacology 1996; 318: 201–204. 9. Balduini, W., Cimino, W. M., Depoortere, H. and Cattabeni, Our experimental data document that thiocolchico- F. Characterization of [3H]thiocolchicoside binding sites in side has a powerful convulsant activity in the rat. The rat spinal cord and cerebral cortex. European Journal of Pharmacology 1999; 376: 149–157. mechanism by which this drug produces focal and 10. Matsubayashi, H., Alkondon, M., Pereira, E. F. R., Swanson, secondarily generalised convulsive status epilepticus K. L. and Albuquerque, E. X. Strychnine: a potent competitive deserves further investigation. Moreover, our findings antagonist of α-Bungarotoxin-sensitive nicotinic acetylcholine confirm the data obtained in humans indicating that the receptors in rat hippocampal neurons. Journal of Pharmacol- ogy and Experimental Therapeutics 1998; 271: 494–506. convulsant activity of thiocolchicoside occurs mainly 11. Grenningloh, G. H., Rienitz, A., Schmitt, B. et al. The when there are minimal lesions of the dura and arach- strychnine-binding subunit of the glycine receptor shows ho- noid membranes. mology with nicotinic acetylcholine receptors. Nature 1987; 328: 215–220. 12. Betz, H. Ligand-gated ion channels in the brain: the amino acid receptor superfamily. Neuron 1990; 5: 383–392. ACKNOWLEDGEMENTS 13. Frostholm, A. and Rotter, A. Glycine receptor distribution in mouse CNS: autoradiographic localization of [3H]strychnine We thank Giuseppe Zucca, Corrado Occhineri and binding sites. Brain Research Bulletin 1985; 15: 473–486. 14. Van den Pol, A. N. and Gorcs, T. Glycine and glycine receptor M. Laura Fois for their assistance. immunoreactivity in brain and spinal cord. The Journal of Neuroscience 1988; 8: 472–492. 15. Olsen, R. W. and DeLorey, T. M. GABA and glycine. In: REFERENCES Basic Neurochemistry: Molecular, Cellular and Medical As- pects (Eds. G. J. Siegel, B. W. Agranoff, R. W. Albers, S. K. Fisher and M. D. Uhler). Philadelphia, Lippincott-Raven, 1. Janbroers, J. M. Review of the toxicology, pharmacodynamics 1999: pp. 336–346. and pharmacokinetics of thiocolchicoside, a GABA-agonist 16. Amato, G., La Grutta, V., Militello, L. and Enia, F. The func- muscle relaxant with anti-inflammatory and analgesic actions. tion of the caudate nucleus in the control of some paroxystic Acta Therapeutica 1987; 13: 221–227. activities in the neuraxis. Archives Internationales de Physi- 2. Dasheiff, R. M. and Rarirez, L. F. The effects of colchicine ologie te de Biochimie 1969; 77: 465–484. in mammalian brain from rodents to Rhesus monkeys. Brain 17. Engel, J. J. Experimental models of epilepsy: classification Research Review 1985; 10: 47–67. and relevance to human epileptic phenomena. In: Neurotrans- 3. Wisniewski, H. and Terry, R. D. Experimental colchicine mitter in Epilepsy (Eds. G. Avanzini, J. J. Engel, R. Fariello encephalopathy. Laboratory Investigation 1967; 17: 577– and U. Heinemann). Amsterdam, Elsevier, 1992: pp. 9–20. 587. 18. Löscher, W. Animal models of epilepsy and epileptic seizures. 4. Sechi, G. P., Rosati, G., Deiana, G. A. et al. Epileptogenic In: Antiepileptic Drugs Pharmacology and Therapeutics (Eds. activity of thiocolchicoside (abstract). Epilepsia 1993; 34 M. J. Eadie and F. J. E. Vajda). Berlin, Springer, 1999: (Suppl. 2): 99–100. pp. 19–62.