Modeling disease. Animal models for genetic disease might help in discovering new treatment options, especially when a large number of drugs or compounds can be tested in this model. In a recent paper in Nature Communications, a zebrafish model for Dravet Syndrome is used for medium-throughput screening of compounds approved by the Foods and Drugs Administration (FDA). The authors identify a single compound that is capable of abolishing behavioral and electrographic seizures in SCN1A-deficient zebrafish.
The zebrafish genome. There is quite some similarity between the zebrafish and the human genome, except that Danio rerio (the zebrafish) has two of them. During evolution, the zebrafish genome duplicated, which makes genetic analysis in zebrafish difficult. The SCN1A gene in zebrafish has two so-called paralogues, namely scn1Laa and scn1lab. During a chemical mutagenesis screen, a zebrafish with a mutation in scn1Lab was identified, which was used for the study by Baraban and collaborators. Given the duplication of the zebrafish SCN1A gene, the authors used animals in which both scn1Lab genes are mutated, generating an animal model that is expected to have ~50% of sodium channels left due to the other remaining paralogue. This is in parallel to Dravet Syndrome, where one copy of the human SCN1A gene is mutated or deleted.
Zebrafish and epilepsy. Baraban and collaborators started by characterizing the scn1Lab mutant zebrafish both through EEG and through behavioral studies. In the EEG, they found that mutant zebrafish had frequent, brief epileptic activity that was not found in control animals. For a behavioral analysis, the authors observed the swimming behavior in the animals. The scn1Lab animals showed whole-body convulsions and undirected swimming behavior. Zebrafish usually stay close to the side of a petri dish, whereas scn1Lab mutants swam randomly through the entire dish. This behavior could be automatically tracked and was simple enough to observe as to allow analysis in a large number of animals. Such relatively easy model systems and the fast generation time of the animals make it possible to perform a large number of experiments, such as medium-throughput testing of various compounds. This assay was used by the authors to investigate known antiepileptic drugs and other compounds.
Medium-throughput screening. The authors first demonstrated the effect of known antiepileptic drugs using EEG in mutant zebrafish. Somewhat similar to the response in patients with Dravet Syndrome, diazepam, bromide, stiripentol and valproate decreased the frequency of epileptic discharges. Ethosuximide, carbamazepine and most prominently vigabatrin increased the frequency of epileptic discharges. When applied to the behavioral test, the swimming behavior was improved by similar antiepileptic drugs, even though some drugs showed paradoxical behavior. For example carbamazepine, earlier shown to increase epileptic discharges, was potent in reducing the pathological swimming behavior. Convinced that the “swimming test” has the potential to test the effect of various compound on seizures in scn1Lab fish, the authors went on to test the effect of 320 new compounds.
Clemizole. Of the 320 compounds, 18 were found to inhibit behavioral seizures. Subsequently, the effect of these compounds on electrographic seizures was tested. A single compound was found to both abolish the behavioral phenotype and electrographic seizures. Clemizole is an H1-antihistamine, i.e. a drug that is used to treat allergic reactions. Clemizole was used in the 1950s and 1960s for that purpose and was reported to be well tolerated. However, it is currently not marketed. The finding by Baraban and collaborators might be a motivation to reinvestigate possible antiepileptic properties of the compound and review the older literature on tolerability and side effects.
Caveats. Even though the paper by Baraban and collaborators provides an interesting way of identifying novel compounds that might warrant further investigation, there are some important caveats when using model systems. Animals can only model certain aspects of the disease, and even though the authors used behavioral and electrographic data, this is limited in its applicability to humans. Other model systems, such as SCN1A-deficient mice, show remarkable response of the autism-like behavior through low dose clonazepam, a phenomenon that is not observed in humans to the same extent. Therefore, it is important to be careful in extrapolating from model systems. However, despite all necessary caution, the study by Baraban and collaborators is the first medium-throughput screen for compounds with a possible effect on Dravet Syndrome. Functional studies like this have the huge potential to transform genetic findings into possible treatment options.