Toward the use of novel alternative methods in epilepsy modeling and drug discovery

“

TL;DR

This review highlights that despite over 20 new antiseizure medications since 1993, one-third of epilepsy patients remain drug-resistant, partly because most treatments only suppress seizures rather than address underlying mechanisms. The authors present alternative models—including organotypic brain slice cultures, iPSC-derived neurons, brain organoids, and larval zebrafish—that can replace, reduce, and refine (3Rs) traditional rodent models for high-throughput drug screening. These novel approach methodologies (NAMs) have already led to clinical successes, such as fenfluramine (FDA-approved) and clemizole (in clinical trials) for Dravet syndrome, demonstrating their translational potential.

Full summary

5,569 CHARS

**Background** Epilepsy is a chronic neurological disorder affecting approximately 50 million people worldwide, characterized by recurrent seizures. Despite the introduction of over 20 novel antiseizure medications (ASMs) since 1993, more than one-third of patients remain resistant to available therapies. This resistance is partly because most existing drugs target seizure suppression rather than the underlying epileptogenic mechanisms. Traditional rodent models have been instrumental in understanding epilepsy pathophysiology and drug discovery, but they are limited by high costs, low throughput, and ethical concerns. The 3Rs principle (replace, reduce, refine) has driven the search for alternative models. This review provides an overview of current gold-standard animal models and explores emerging alternatives, focusing on ex vivo, in vitro, and in vivo larval zebrafish models that contribute to the 3Rs in epilepsy modeling and drug screening.

**Methods** The authors conducted a narrative review of the literature, summarizing the state of the art in epilepsy modeling. They describe pharmacological and genetic methods currently available, including electrical and chemical seizure induction in rodents (e.g., maximal electroshock seizure test, 6-Hz psychomotor seizure model, pentylenetetrazole (PTZ) test, pilocarpine and kainate models) and genetic animal models (e.g., mouse models of lissencephaly and Dravet syndrome). The review then details alternative models: organotypic brain slice cultures (OSCs), induced pluripotent stem cells (iPSCs), brain organoids, and larval zebrafish (Danio rerio). For zebrafish, both pharmacological models (PTZ and kainic acid) and genetic models (e.g., scn1Lab mutant) are described, including the use of CRISPR/Cas9-based gene editing (CRISPANTs, base editing, prime editing) for high-throughput disease modeling.

**Key Results** - Rodent models: The PTZ test in rodents has been crucial for identifying many clinically used ASMs. In the 6-Hz psychomotor seizure model, repeated corneal stimulation in mice established a kindling model resistant to ASMs. The amygdala kindling rat model of temporal lobe epilepsy (TLE) is a well-established pharmacoresistant model. - Organotypic brain slice cultures (OSCs): OSCs preserve three-dimensional structure and synaptic organization, allowing long-term live imaging and electrophysiology. They can be derived from human tissue obtained from neurosurgery, preserving human pyramidal neuron properties. However, slice preparation affects viability and neuronal connections. - iPSCs: Patient-specific iPSC-derived neurons have been used to model Dravet syndrome (SCN1A mutation), showing loss of GABAergic inhibition. CRISPR/Cas9 in iPSCs has generated isogenic lines for KCNQ2 encephalopathy, revealing functional enhancement of Ca²⁺-activated K⁺ channels. Limitations include variable expression profiles and difficulty recapitulating brain complexity. - Brain organoids: Organoids from EPM1 patients (CSTB mutation) showed altered progenitor proliferation and premature differentiation. UGDH mutant organoids were reduced in size with decreased neuronal progenitor markers, while UGDH mutant zebrafish did not show the same defects, highlighting model-specific differences. Organoids have also modeled Angelman syndrome and Rett syndrome, with hyperactive neuronal firing and epileptiform spikes. - Zebrafish models: The PTZ model in zebrafish larvae (7 days post-fertilization) exhibits electrophysiological, behavioral, and molecular changes similar to rodent PTZ models. ASMs such as valproic acid (300 μM–10 mM), carbamazepine (10–100 μM), diazepam (30–100 μM), gabapentin (1–10 mM), and lacosamide (100 μM–3 mM) showed concentration-dependent increases in seizure latency, while pregabalin did not. The scn1Lab mutant zebrafish recapitulates Dravet syndrome phenotypes, and phenotype-based screening of over 300 compounds identified clemizole (EPX-100) as effective, which has passed phase I clinical trials. Fenfluramine (now FDA-approved as Fintepla®) and other repurposed drugs (trazodone, lorcaserin) were also identified. A novel kainic acid model via intrapericardial injection in 3 dpf larvae induced epileptiform discharges, reduced by topiramate 100 μM, tiagabine 100 μM, and carbamazepine 100 μM, but none decreased seizure-like behavior. - Genetic models: CRISPR/Cas9-generated CRISPANTs for scn1lab showed behavioral fingerprints highly correlated with stable mutants. Base editing and prime editing have achieved up to 30% correct edits in F0 zebrafish embryos, enabling rapid generation of humanized models.

**Clinical Implications** The integration of NAMs into epilepsy research offers a path to more ethical, cost-effective, and high-throughput drug discovery. The success of zebrafish models in identifying clemizole and fenfluramine for Dravet syndrome demonstrates translational potential. The FDA Modernization Act 2.0 (2022) now allows alternative methods to replace animal testing for drug approval. A combinatorial approach, such as Integrated Approaches for Testing and Assessment (IATA), could combine data from multiple NAMs to evaluate compound efficacy, reducing and eventually replacing animal use. Advances in genome editing (CRISPR/Cas9, base editing, prime editing) enable personalized treatment strategies by modeling patient-specific mutations. Over the next 5 years, cell-based and larval models are expected to grow, accelerating the discovery of disease-modifying therapies for epilepsy and other neurological disorders.