Antiepileptic natural product discovery using a zebrafish seizure model

Epilepsy is a common neurological disorder marked by the occurrence of spontaneous seizures caused by abnormal neuronal activity in the brain. About 65 million people worldwide are estimated to have epilepsy. Many epilepsy patients can be treated effectively with currently available anti-epileptic drugs, but for 25-30% of this patient population, adequate seizure control is not achieved. Even if patients can undergo surgery, only 45-75% stays seizure free. For thesereasons, the discovery of novel anticonvulsant lead compounds and the subsequent development of new, effective antiepileptic drugs (AEDs) remains an important area of research.Zebrafish offer several advantages that make them an attractive in vivo model for different diseases. First of all they show a high genetic, physiological and pharmacological similarity to humans. Additionally, the embryos and larvae develop rapidly ex utero and are optical transparent, making detection of morphological and behavioral effects possible. Recently zebrafish have also emerged as in vivo model for epilepsy. For instance, seven-day-old zebrafish larvae exposed to the GABAA antagonist pentylenetetrazol (PTZ) exhibit increased locomotor activity, seizure-like behavior, and epileptiform electrographic activity. A previous study showed that 12 out of 13 AEDs suppressed PTZ-mediated increases in larval movement, indicating the potential utility of zebrafish as a high-throughput in vivo model for AED discovery. However, a question remained as to whether an AED-induced decrease in locomotion is trulyindicative of anticonvulsant activity, as some drugs may impair larval movement through other mechanisms such as general toxicity or sedation. In order to validate further the model, we carried out a study in PTZ-treated zebrafish larvae, to directly compare the ability of AEDs to inhibit seizure-like behavioral manifestations with their capacity to suppress epileptiform electrographic activity. We re-tested the 13 AEDs of which 12 were previously reported to inhibit convulsions in the larval movement tracking assay, administering concentrations that did not, on their own, impair locomotion. In parallel, we carried out open-field recordings on larval brains after treatment with each AED. For 10 out of 13 AEDs we obtained the same response in both the behavioral and electrographic assays. These AEDs are: carbamazepine, diazepam, ethosuximide, gabapentin, lamotrigine, levetiracetam, oxcarbazepine, phenytoin, primidone and sodium valproate. Tiagabine, topiramate and zonisamide gave a different response when the behavioral and electrographic assays were compared. Topiramate and zonisamide were positive in the behavioral assay and negative in the electrographic assay, whereas for tiagabine it was the other way round. Only diazepam, ethosuximide and sodium valproate tested positive in bothassays.Overall our data correlate well with those reported in the literature for acute rodent PTZ tests, indicating that the larval zebrafish brain is more discriminatory than previously thought in its response to AEDs with different modes of action.Our results underscore the validity of using the zebrafish larval locomotor assay as a rapid first-pass screening tool in assessing the anticonvulsant and/or proconvulsant activity of compounds, but also highlight the importance of performing adequate validation when using in vivo models.Subsequent, this validated behavioral zebrafish method was used in the screening ofmedicinal plants in order to discover possible new AEDs.Medicinal plants are still used in many developing countries in the treatment of various diseases instead of modern medicine. One of the indications is epilepsy. The majority of the patient rely on the treatment given by traditional healers. Many different plants are used for the treatment of epilepsy, dependent on the region.In Eastern Congo a decoction of 14 plants is used by traditional healers for the treatment of patients with seizures. In collaboration with a neurologist of the University of Bukavu, 10 of these plants were identified, collected and extracted for screening in the behavioral zebrafish seizure model. An extract of the leaves of Indigofera arrecta, a large shrub belonging to the family of Fabaceae and locally called Kasholoza , particularly showed potent anticonvulsant activity. Hence, we set out to isolate and identify the secondary bio-active metabolites of the plant. Bioassay-guided fractionation using zebrafish as an in vivo model for epilepsy and structural using high-resolution mass spectrometry and NMR resulted in the isolation of indirubin.Purified indirubin was used for further testing, showing a positive effect in the behavioral andin the electrographic assay in zebrafish. Indirubin also showed antiepileptic activity in rodent models, in the 6 Hz model in mice and in the intrahippocampal pilocarpine limbic seizure model inrats. Testing a number of compounds with different affinities for the primary targets (GSK-3ß, CDK andAhR) of indirubin in the zebrafish behavior assay showed us that the anticonvulsant activity of indirubin was probably due to GSK-3ß inhibition.Therefore, two other selective GSK-3ß inhibitiors were chosen for testing in a panel of zebrafish and rodent seizure models. BIO-acetoxime, a potent indirubin-like GSK-3alfa/ß inhibitor, and TCS2002, a potent non-indirubin related GSK-3ß inhibitor, also proved to have anticonvulsant activity in almost the same set of experiments as indirubin. Only TCS2002 gave a different result in the electrographic zebrafish assay.In addition, targeted inhibition of GSK-3ß using antisense-mediated morpholino knockdown wasable to counter PTZ-induced seizure behavior in a concentration dependentmanner in zebrafish.Based on this study indirubin, BIO-acetoxime and TCS2002 may be possible new AEDs with a new mechanism of action. Our results strongly point to GSK-3ß inhibition as being responsible for the antiepileptic effect. Therefore, GSK-3ß inhibitors may represent a potential new class of AEDs with a novel mechanism of action.nrpages: 125status: publishe