kcnb1 loss of function in zebrafish causes neurodevelopmental and epileptic disorders associated with γ‐aminobutyric acid dysregulation

Abstract

Objective

KCNB1 encodes an α‐subunit of the delayed‐rectifier voltage‐dependent potassium channel K_v2.1. De novo pathogenic variants of KCNB1 have been linked to developmental and epileptic encephalopathies (DEEs), diagnosed in early childhood and sharing limited treatment options. Loss of function (LOF) of KCNB1 has been proposed as the pathogenic mechanism in these disorders. Here, we aim to characterize a knockout zebrafish line targeting kcnb1 (kcnb1 ^+/− and kcnb1 ^−/−) for investigating DEEs.

Methods

This study presents the phenotypic analysis of a kcnb1 knockout zebrafish model, obtained by CRISPR/Cas9 (Clustered Regularly Interspaced Short Palindromic Repeats) mutagenesis. Through a combination of immunohistochemistry, behavioral assays, electrophysiological recordings, and neurotransmitter quantifications, we have characterized the expression, function, and impact of this kcnb1 LOF model at early stages of development.

Results

In wild‐type (WT) larval zebrafish, kcnb1 was found in various regions of the central nervous system and in diverse cell subtypes including neurons, oligodendrocytes, and microglial cells. Both kcnb1 ^+/− and kcnb1 ^−/− zebrafish displayed impaired swimming behavior and “epilepsy‐like” features that persisted through embryonic and larval development, with variable severity, which was restored by the human K _v 2.1 WT DNA. When exposed to the chemoconvulsant pentylenetetrazol (PTZ), both knockout models showed elevated locomotor activity. In addition, PTZ‐exposed kcnb1 ^−/− larvae exhibited increased bdnf mRNA expression and higher c‐Fos fluorescence intensity in cells of the telencephalon. This same model presents spontaneous and provoked epileptiform‐like electrographic activity associated with γ‐aminobutyric acid dysregulation, whereas the brain anatomy and neuronal circuit organization remained unaffected.

Significance

We conclude that kcnb1 knockout in zebrafish leads to early onset phenotypic features reminiscent of DEEs, affecting neuronal functions and primarily inhibitory pathways in developing embryonic and larval brains. This study highlights the relevance of this model for investigating developmental neuronal signaling pathways in KCNB1‐related DEEs.

Key points.

• kcnb1 is expressed in specific cell subtypes and in various regions of the central nervous system in WT larval zebrafish.

• Brain anatomy and neuronal circuits are not disrupted in the kcnb1 lossoffunction zebrafish model.

• Loss of kcnb1 leads to altered behavioral phenotype, and light‐ and sound‐induced locomotor impairments.

• kcnb1 knockout zebrafish larvae exhibit elevated locomotor sensitivity to PTZ and increased expression of epileptogenesis‐related genes.

• kcnb1 ^−/− larvae show spontaneous and provoked epileptiformlike electrographic activity associated with γ‐aminobutyric acid dysregulation

1. INTRODUCTION

KCNB1 encodes the pore‐forming α1 subunit of the voltage‐gated potassium channel subfamily 2 (K_v2.1). K_v2.1 is extensively expressed across the central nervous system (CNS) and predominantly localized in clusters on neuronal soma, proximal dendrites, and axonal initial segment of neurons, ¹ , ² , ³ , ⁴ generating a delayed‐rectifier outward potassium current. ¹ , ² , ⁵ , ⁶ Moreover, the channel modulates neuronal excitability through activity‐dependent regulation of K_v2.1 phosphorylation in various neuronal subtypes. ¹ , ² , ⁵ , ⁶ Beyond its role in ionic conduction, K_v2.1 channels participate in intracellular protein trafficking and calcium signaling. ⁵ , ⁷ , ⁸

Numerous heterozygous de novo pathogenic variants of KCNB1 have been reported in patients with developmental and epileptic encephalopathy (DEE), as well as developmental encephalopathy without epilepsy or with late onset, severe, and pharmacoresponsive epilepsy. ⁹ , ¹⁰ , ¹¹ , ¹² Starting within the first year of life, all patients reported exhibit poor long‐term outcome associated with a wide phenotypic spectrum including severe intellectual disability, attention disorders, and autism spectrum disorder. ⁹ , ¹² , ¹³ The epileptic manifestations encompass a wide range of seizure types and severities, including focal, generalized tonic–clonic, myoclonic–atonic, infantile spasms, and absence seizures. ⁹ , ¹² , ¹⁴ , ¹⁵

In vitro studies of KCNB1 variants revealed various degrees of loss of function (LOF) including reduced potassium conductance, altered ion selectivity, and diminished K_v2.1 channel expression at the cell surface. ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ Some mutations exhibited a dominant‐negative effect when coexpressed with the wild‐type (WT) variant. ¹⁶ , ²⁰ , ²¹ The findings are supported by rodent models, including Kcnb1 knockout (KO) and knockin mice. These mouse models exhibit altered behaviors such as locomotor hyperactivity, anxietylike behavior, lower seizure thresholds when exposed to chemoconvulsants, and spontaneous and/or provoked seizures depending on the model. ²⁰ , ²² , ²³ , ²⁴ , ²⁵

The underlying pathogenic mechanisms of KCNB1‐related DEEs remain largely uncharacterized in rodent models, highlighting the need for alternative genetic models. Zebrafish have gained prominence for studying human neurological disorders due to their genetic manipulability, rapid development, and suitability for high‐throughput drug screening. ²⁶ , ²⁷ , ²⁸ , ²⁹ Larval zebrafish, when exposed to the proconvulsant pentylenetetrazol (PTZ), manifest behaviors akin to seizures, making them a valuable model for epilepsy investigations ³⁰ such as Dravet syndrome related to SCN1A ³⁰ , ³¹ , ³² and epilepsies associated with DEPDC5. ³³ , ³⁴ In addition, drug screening in larval zebrafish has been performed in both genetic and chemically induced epilepsy models, demonstrating the translational potential of this model by identifying compounds capable of reducing the severity of epileptic seizures. ³⁵ , ³⁶ , ³⁷ , ³⁸

The zebrafish ortholog of KCNB1, kcnb1, shares 67% sequence homology with the human gene. Its expression begins at 19 hours post fertilization (hpf) and is primarily localized in the brain of adult zebrafish with tissue‐specific expression patterns comparable to mammals. ³⁹ The first kcnb1 KO zebrafish model (kcnb1 ^−/−) was generated using CRISPR/Cas9 (Clustered Regularly Interspaced Short Palindromic Repeats) mutagenesis, which led to a premature stop codon. ⁴⁰ This model revealed gastrulation defects and reduced brain ventricular system in a subset of embryos, and disruptions of inner ear development leading to balance and hearing deficits. ³⁹ , ⁴⁰ Co‐expression of the human K _v 2.1 with the zebrafish silent kcng4b subunit in HEK cells or Xenopus oocytes modulates the biophysiological properties of K_v2.1. ⁴⁰ , ⁴¹ kcng4b constructs led to a reduced potassium current density (WT and C‐terminal truncated constructs) and to a dominant‐negative effect (N‐terminal truncated and ectopic transmembrane domain constructs). ⁴⁰ , ⁴¹ These findings emphasize the essential role of kcnb1 in developmental signaling pathways and its modulation by silent subunits.

However, the effects of kcnb1 LOF on brain in development, in the context of epilepsy, remain unexplored. In this study, we aim to characterize the behavioral, electrophysiological, and molecular consequences of this kcnb1 ^−/− zebrafish model at various early developmental stages. This model presents the key features of KCNB1‐related DEEs, contributing to a better understanding of the pathophysiological mechanisms underlying this disorder.

2. MATERIALS AND METHODS

2.1. Housing conditions of zebrafish

All procedures were approved by the institutional ethics committees at the research centers of the Imagine Institute (INSERM U1163, Paris, France) and were performed in accordance with the European Union Directive (2010/63/EU). Adult zebrafish (Danio Rerio) were housed in a conventional animal facility. Experiments were performed on WT AB and TU strains named kcnb1 ^+/+ and kcnb1 KO zebrafish lines (kcnb1 ^+/− and kcnb1 ^−/−) staged from 0 to 6 days post fertilization (dpf). ⁴² Fertilized eggs were collected by natural spawning. Embryos and larvae were maintained at 28 ± 1°C in a non‐CO₂ incubator (VWR) with a 14‐h/10‐h light/dark cycle in embryo medium (Instant Ocean).

2.2. kcnb1 KO zebrafish lines

Heterozygous kcnb1 KO embryos (kcnb1 ^+/−) were provided by Dr. Vladimir Korzh (Warsaw, Poland). The kcnb1 KO line was generated using CRISPR/Cas9 mutagenesis, resulting in a premature stop codon. ⁴⁰ kcnb1 ^+/− KO embryos were raised to adulthood and crossed to produce kcnb1 ^−/− zebrafish for experiments in our animal facility (kcnb1 ^sq301/sq301 , ZDB‐ALT‐170417‐2). ⁴⁰ Genomic DNA was extracted from adult zebrafish fins. Samples were amplified by polymerase chain reaction (PCR; Bio‐Rad Laboratories) using DreamTaq Hot Start PCR Master Mix (Thermo Fisher Scientific) and primers targeting kcnb1 (10 mmol·L⁻¹): F_5′‐TGTGACGACTACAACCTGGA‐3′ and R_5′‐CTCCTCGTTCATCTGCTCCT‐3′. DNA samples were sent for sequencing to GATC Biotech, Eurofins Genomics.

2.3. Survival assay and morphological analysis

Zebrafish embryos and larvae were maintained at 28 ± 1°C in a non‐CO₂ incubator (VWR) and were fed daily from 0 to 15 dpf. After 6 dpf, larvae were transferred to tanks with dripped water flux. Survival was assessed based on the number of living individuals relative to the total population. Zebrafish were photographed using a stereomicroscope (SZX16, Olympus Life Science). Measurements of the body length and the head surface were performed manually with ImageJ software (National Institutes of Health).

2.4. Reverse transcription quantitative PCR

cDNA from larval zebrafish at 6 dpf (30 per pool) was synthesized using 5X All‐In‐One RT MasterMix (abm). Quantitative PCR (qPCR) on kcnb1 and bdnf was performed with BlasTaq 2X qPCR Master Mix (abm) on a CFX384 System (Bio‐Rad Laboratories). Relative gene expression was determined by the 2−^ΔΔCt method, normalized to β‐actin or ef1α, with kcnb1 ^+/+ serving as the reference (relative fold change = 1). The following primers were used for qPCR: β‐actin (F_5′‐CGAGCTGTCTTCCCATCCA‐3′, R_5′‐TCACCAACGTAGCTGTCTTTCTG‐3′), ef1α (F_5′‐CTGGAGGCCAGCTCAAACAT‐3′, R_5′‐ATCAAGAAGAGTAGTACCGCTAGCATTAC‐3′), kcnb1 (F_5′‐TGAAGTTCCGGGAGAGTGTT‐3′, R_5′‐CAGGTTGGCGATGTCGTTCT‐3′), and bdnf (F_5′‐GACTCGAAGGACGTTGACCTGTA‐3′, R_5′‐CGGCTCCAAAGGCACTTG‐3′).

2.5. Locomotion assessment

2.5.1. Premotor activity

Tail‐coiling activity of embryos was recorded every hour from 24 to 36 hpf. Sixty‐second videos were obtained under a stereomicroscope with darkfield illumination (SZX16, Olympus Life Science), recorded at 30 frames per second (fps). Premotor activity was counted manually using ImageJ software.

2.5.2. Touch‐evoked escape response

The tails of 48‐hpf embryos were mechanically stimulated, and their swimming trajectories were recorded at 30 fps using SpinView V2.0.0.147 software (FLIR Systems). The trajectory was traced using the Manual Tracking Plug‐in in ImageJ software, and analyzed for distance, velocity, and time spent in motion (see Figure 3A). Microinjection of a linearized plasmid encoding the human K _v 2.1‐WT DNA (hK _v 2.1‐WT, Invitrogen) was conducted in one‐cell stage eggs from kcnb1 ^+/− and kcnb1 ^−/− lines. The optimal concentration at 150 ng/μL ensured minimal morphological deformities and low mortality rates. The rescue of the locomotor phenotype was evaluated using the touch‐evoked escape response (TEER) test.

FIGURE 3.

Loss of kcnb1 leads to an altered behavioral phenotype, and light‐ and sound‐induced locomotor impairments. (A) Schematic representation of the touch‐evoked escape response (TEER) test. The tail of a 48 hours post‐fertilization (hpf) embryo is mechanically stimulated, and the swimming trajectory of zebrafish is recorded. (B) Top part: Representative traces of individual swimming episodes at 48 hpf showing the typical tortuous trajectory of kcnb1 knockout models (kcnb1 ^+/− and kcnb1 ^−/−) as compared to the straight‐line trajectory of kcnb1 ^+/+ zebrafish (5 trajectories/genotype). Bottom part: One‐cell stage eggs from kcnb1 ^+/− and kcnb1 ^−/− zebrafish were microinjected with the hK _v 2.1 wild‐type (WT) plasmid, and the trajectory of embryos was analyzed at 48 hpf using the TEER test. The expression of the hKv2.1‐WT construct allowed the rescue of the locomotor phenotype of kcnb1 knockout models by reproducing a more linear trajectory similar to WT embryos. (C–E) The quantification of the swimming trajectory tortuosity shows a significant increase of different parameters studied in both kcnb1 knockout conditions as compared to kcnb1 ^+/+ embryos, including an increase in (C) the distance swam, (D) the velocity, and (E) the time spent in motion. Furthermore, within the same genotype, a huge variability in the swimming behavior has been observed, and zebrafish were divided into two distinct phenotypes: severe (at least two swim circles) and mild (other trajectories). The subdivision of both phenotypes is represented in Table S2 and Figure S2C–F (N = 3; n = 50–88/genotype; one‐way analysis of variance [ANOVA] with Bonferroni post hoc test). (F) Distance swam by embryos at 48 hpf following a rescue with the hKv2.1‐WT DNA construct and following a TEER test. Results show a significant reduction of the distance traveled by kcnb1 knockout models due to the expression of the hKv2.1‐WT DNA (N = 3; n = 10–37/condition; unpaired t‐test). (G) Representation of the average distance traveled by larvae at different days of development (from 3 to 6 days post‐fertilization [dpf]) following a 10‐min light–dark protocol, each period repeated 3 times. At 3 and 4 dpf, kcnb1 ^+/+ and kcnb1 knockout zebrafish (kcnb1 ^+/− and kcnb1 ^−/−) presented low locomotor activity that was significantly increased from 5 to 6 dpf. A significant gap was observed starting from 5 dpf with locomotor hyperactivity for the kcnb1 ^−/− condition as compared to kcnb1 ^+/+ and kcnb1 ^+/− zebrafish, reflecting the result obtained in Figure 3I. Colored statistical markers represent the comparison of locomotor activity of zebrafish with the same genotype, using the reference value at 3 dpf. Black statistical markers indicate comparisons between the three genotypes at specific time points (N = 3; n = 35/genotype; one‐way ANOVA with Bonferroni post hoc test). (H) Schematic representation of individual trajectory of three zebrafish larvae per genotype obtained with ViewPoint software (Zebrabox). Between 3 and 6 dpf, a protocol was applied to larvae with three repetitions of 10 min in the light followed by 10 min in the dark (green lines: slow movements, <8 mm/s; red lines: fast movements, >8 mm/s). (I) Average distance swam by zebrafish at 6 dpf during the whole light–dark protocol described previously. Locomotor hyperactivity was observed in kcnb1 ^−/− larvae in the light and continued to be significantly increased during the dark phase as compared to kcnb1 ^+/+ and kcnb1 ^+/− zebrafish (N = 3; n = 35/genotype; one‐way ANOVA with Bonferroni post hoc test). (J) Quantification of the distance swam during the 5 s following each audio stimuli applied (450 Hz, 80 dB, 1 s) to larvae at 6 dpf. Both knockout conditions (kcnb1 ^+/− and kcnb1 ^−/−) present a significant decrease of the locomotor activity in response to audio stimuli as compared to the kcnb1 ^+/+ condition (N = 3; n = 48–64/genotype; one‐way ANOVA with Bonferroni post hoc test). *p < .05, **p < .01, ***p < .001, ****p < .0001. ns, non‐significant.

2.5.3. Spontaneous locomotion and PTZ‐induced seizures

Larvae were acclimatized in a 48‐well plate (Thermo Fisher Scientific) with lights off for 15 min into a Zebrabox equipped with Zebralab 438 software (Viewpoint Life Sciences). Larvae were submitted to two different protocols: (1) a light‐induced protocol of 60 min recording split into 10 min light/dark conditions repeated three times and (2) a sound‐induced protocol of four repeated audio stimuli (450 Hz, 80 dB, 1 s/audio). The 5 s following audio stimuli was analyzed. To evaluate the effect of PTZ on locomotor activity, larvae were first recorded for 30 min to determine baseline activity levels. Fresh 5 mmol·L⁻¹ PTZ (Sigma‐Aldrich) was then added, and zebrafish were recorded for a further 30 min (see Figure 4A).

FIGURE 4.

kcnb1 knockout zebrafish exhibit increased locomotor sensitivity to pentylenetetrazol (PTZ) associated with an increased expression of epileptogenesis‐related genes. (A) Schematic representation of the protocol followed to identify the impact of the proconvulsant PTZ on locomotion, and bdnf and c‐Fos expression. To test the seizure susceptibility in the kcnb1 knockout model, we recorded the baseline locomotor activity of 6 days post‐fertilization (dpf) larvae for 30 min, followed by a 30‐min exposure to 5 mmol·L⁻¹ PTZ (Zebrabox, ViewPoint Life Sciences). We quantified the expression of two epileptogenesis‐related genes at the end of each basal and provoked‐seizure periods: bdnf (brain‐derived neurotrophic factor) by quantitative polymerase chain reaction (qPCR) and c‐Fos by immunohistochemistry (IF). GABA, γ‐aminobutyric acid. (B) Schematic representation of individual trajectory of three zebrafish larvae per genotype obtained with ViewPoint software (Zebrabox). Fast swimming circles were observed for both mutant larvae conditions (kcnb1 ^+/− and kcnb1 ^−/−) after 30 min of 5mmol·L⁻¹ PTZ treatment. Green lines: slow movements (<8 mm/s); red lines: fast movements (>8 mm/s). (C) Global locomotor activity of 6‐dpf zebrafish in the dark was recorded by applying a protocol of 30 min of basal activity followed by 30 min of chemically induced seizures using PTZ treatment at 5 mmol·L⁻¹. Both kcnb1 ^+/− and kcnb1 ^−/− larvae showed a significant increase in distance traveled during the 30‐min recording after the chemical treatment as compared to kcnb1 ^+/+ zebrafish (N = 3; n = 27/genotype; one‐way analysis of variance [ANOVA] with Bonferroni post hoc test; +/− and −/− vs. +/+ after PTZ treatment: **p < .01; ****p < .0001). (D) Reverse transcription qPCR analysis of total bdnf mRNA at 6 dpf, an epileptogenesis‐related gene, before and after 5mmol·L⁻¹ PTZ treatment (N = 3; n = 30/sample; one‐way ANOVA with Bonferroni post hoc test; ****p < .0001). Data are normalized to ef1α mRNA expression, and kcnb1 ^+/+ nontreated larvae are considered as the reference value (relative fold change = 1). The kcnb1 ^−/− condition presents a tendency toward an increased bdnf transcript expression during the basal locomotor activity, which is confirmed by a significantly increased expression after chemically induced‐seizures as compared to kcnb1 ^+/+ and kcnb1 ^+/−. (E) Whole‐mount images of 6‐dpf zebrafish immunostained with anti‐c‐Fos, an immediate–early neuronal and epileptogenesis‐related gene, obtained 30 min after basal activity and 5mmol·L⁻¹ PTZ treatment periods. The telencephalon (T) was the major region activated after the chemical treatment for each genotype (n = 8–11/condition, dorsal view, three‐dimensional reconstruction, scale bar = 50 μm, magnification = 20×). (F) Quantification of the number of c‐Fos‐positive neurons (nb) normalized to the volume (μm³) of the telencephalon of 6‐dpf larvae during their basal or chemically treated periods using IMARIS v10.1.0 software (Oxford Instruments; see Figure S3A). The results did not show any difference in the number of activated neurons in both knockout conditions (kcnb1 ^+/− and kcnb1 ^−/−) as compared to the kcnb1 ^+/+ condition in pre‐ and post‐PTZ treatment periods (n = 8–11/condition; Mann–Whitney test). (G) Distribution (as a percentage) of c‐Fos‐positive neurons according to the fluorescence intensity value of c‐Fos in the telencephalon of zebrafish at 6 dpf, reflecting the level of neural activation that was divided into four equal shares (from 0% to 100%): low (0%–25%), moderately low (25%–50%), moderately high (50%–75%), and high (75%–100%) neural activation. The three nontreated conditions presented a similar distribution, with a majority of low neuronal activation of c‐Fos positive neurons. However, the chemically treated kcnb1 ^−/− zebrafish presented higher global activation, with a significant shift to moderately low and moderately high neuronal activation (see Figure S3B). n = 8–11/condition; Mann–Whitney test; *p < .05. ns, non‐significant.

2.6. Electrophysiological analysis

Larval zebrafish at 5 and 6 dpf were embedded in 1% low‐melting‐point agarose (Sigma‐Aldrich) and covered with artificial cerebrospinal fluid (ACSF; pH 7.8, osmolarity to 290–295 mOsm/L). A microelectrode (2–7 MΩ) was filled with ACSF and implanted into the optic tectum, using an Olympus microscope (Infinity 3S, 10× magnification). Local field potential (LFP) recordings were obtained using a MultiClamp700B amplifier (Molecular Devices) coupled with an Axon Digidata 1550 (Molecular Devices) and Clampex V11.1 software (Molecular Devices). Baseline recordings of 30‐min duration were performed in larvae before adding 40 mmol·L⁻¹ PTZ (Sigma‐Aldrich). After 5 min of treatment, the recording continued for 30 min. Negative spikes were automatically analyzed using Clampfit V11.2 software (Molecular Devices) during a 10‐min interval, using a low‐pass Gaussian filter of 560 Hz and a digital reduction of 10 (see Figure 5A,B).

FIGURE 5.

kcnb1 ^−/− larvae show spontaneous and provoked epileptiform‐like electrographic activity associated with a dysregulation of γ‐aminobutyric acid (GABA). (A, B) Schematic representation of the protocol applied for electroencephalographic recordings and neurotransmitter quantification of 5 and 6 days post‐fertilization (dpf) zebrafish. Neuronal activity in the optic tectum of zebrafish was recorded by applying a protocol of 30 min of basal activity followed by 30 min of 40mmol·L⁻¹ pentylenetetrazol (PTZ) treatment. A 10‐min segment in the middle of each period (basal activity and provoked seizures) was used to analyze the total number of negative spikes and duration of events for each genotype. Enzyme‐linked immunosorbent assay (ELISA) was performed to quantify GABA and glutamate following a 30‐min pre‐ and post‐PTZ treatment at 5 mmol·L⁻¹. (C; a) Representative traces of electroencephalographic recordings in the optic tectum of kcnb1 ^+/+ and kcnb1 knockout (kcnb1 ^+/− and kcnb1 ^−/−) larvae showing spontaneous and provoked epileptiform‐like electrographic activity in the kcnb1 ^−/− model characterized by (b) interictal‐like activity, (c) ictal‐like activity, (d) polyspike discharges, and (e) large‐amplitude spikes. (D, E) Quantification of electrophysiological recordings by analyzing (D) the total number of negative spikes and (E) the duration of events, over 10 min as described in panel A. kcnb1 ^−/− larvae showed significantly increased spontaneous and provoked neuronal activity, reflected by a significant elevation of the number of spikes, although the duration of events was similar to the kcnb1 ^+/+ condition in pre‐ and post‐PTZ treatment. kcnb1 ^+/− larvae presented a profile similar to the kcnb1 ^+/+ condition in terms of number of spikes but showed a significant increase of event duration (see Table S3; n = 3–5/genotype; unpaired t‐test). (F, G) Quantification of (F) GABA and (G) glutamate by ELISAs in the head of 6‐dpf larvae following the 30‐min pre‐ and post‐5mmol·L⁻¹ PTZ treatment. kcnb1 ^−/− zebrafish present significantly increased GABA levels observed in both baseline and post‐PTZ conditions conversely to kcnb1 ^+/− larvae and compared to the kcnb1 ^+/+ condition. We observed a lack of significant changes in glutamate levels in both kcnb1 knockout models (kcnb1 ^+/− and kcnb1 ^−/−) in pre‐ and post‐PTZ conditions (N = 3–5; n = 50/sample; unpaired t‐test). For panels D–G, colored statistic indications correspond to mutant conditions compared to the kcnb1 ^+/+ condition in pre‐ or post‐PTZ treatment. *p < .05, **p < .01, ***p < .001, ****p < .0001. ACSF, artificial cerebrospinal fluid; ns, non‐significant.

2.7. Immunohistochemistry on slices

Zebrafish aged from 48 hpf to 6 dpf were fixed as previously described. ³⁴ Brain slices of zebrafish (20 μm) were obtained using the cryostat CM3050S (Leica). After permeabilization for 30 min, the slices were incubated overnight at 4°C in blocking solution containing primary antibodies: Kcnb1 (1:100, Tebubio, #PAB7569), NeuN (1:100, Merck, #ABN90), Olig2 (1:100, DSHB, #PCRP‐OLIG2‐1E9‐s), CX3CR1 (1:100, Proteintech, #13885‐1‐AP), Ankyrin G (1:100, Proteintech, #27980‐1‐AP), and HuC (1:100, Tebubio, #FNab04072). Secondary antibodies (Alexa Fluor 488, Alexa Fluor 568, and Alexa Fluor 647, Thermo Fisher Scientific) were incubated in blocking solution for 2 h at room temperature (RT). Slices were incubated in 4,6‐diamidino‐2‐phenylindole (Invitrogen, #D3571) followed by mounting on glass slides in Immu‐Mount mounting medium (Epredia, #9990402). Images were captured using a spinning disk Zeiss system (Carl Zeiss). The colocalization, indicated in white, was determined using Z‐stack projection on IMARIS v10.1.0 software (Oxford Instruments).

2.8. Whole‐mount immunohistochemistry

Whole‐mount immunohistochemistry on 48 hpf and 6 dpf zebrafish was performed according to a previously established protocol. ⁴³ Embryos were treated with .003% 1‐phenyl‐2‐thiourea (Sigma‐Aldrich) to prevent pigmentation of the skin. Zebrafish were fixed with 4% formaldehyde (Sigma‐Aldrich) for 2 h at RT. Dehydration with methanol (Sigma‐Aldrich) was followed by gradual rehydration. Zebrafish were permeabilized in fresh acetone (Sigma‐Aldrich) for 20 min and blocked with 10% bovine standard albumin (BSA; Eurobio Scientific) diluted in Phosphate Buffered Saline and .1% Tween solution (.1% PBST) overnight at 4°C. They were incubated with the primary antibody (1% BSA/.1% PBST) at 4°C for two days. Primary antibodies used at 1:100 were as follows: acetylated tubulin (Sigma‐Aldrich, #T7451), 3A10 (DSHB, #3A10) and c‐Fos (Santa Cruz, #sc‐166 940). Secondary antibodies were incubated in 1% BSA/.1% PBST overnight at 4°C using Alexa Fluor 488 and Alexa Fluor 647 (Thermo Fisher Scientific) at 1:250. After passage through increased percentages of glycerol (Sigma‐Aldrich), image acquisition was performed under a spinning disk Zeiss system (Carl Zeiss). The same parameters were applied to all images using ImageJ.

2.9. γ‐Aminobutyric and glutamate enzyme‐linked immunosorbent assays

The heads of 6‐dpf larvae (50 per pool) were collected at the end of each 30‐min recording of the basal activity and 5 mmol·L⁻¹ PTZ treatment (see Figure 5A). The same reagents were used as the IP protocol for tissue homogenizing. Enzyme‐linked immunosorbent assay was performed to quantify γ‐aminobutyric acid (GABA; ImmuSmol, #BA E‐2500) and glutamate (ImmuSmol, #BA E‐2400) from 20 μg of supernatant, following the manufacturer's protocol.

2.10. Statistical analysis

Statistical analysis was performed using Prism v8 software (GraphPad Software). Statistical tests used are specified in the legend of each figure. Values represent the mean ± SEM. Results were considered significant when the p‐value was <.05. For all experiments, at least three single experiments were performed (N) with a certain number of embryos or larvae per condition (n).

3. RESULTS

3.1. kcnb1 is expressed in diverse cell subtypes and regions of the CNS in WT larval zebrafish

Previous studies have shown significant expression of kcnb1 in the eyes, ears, and CNS of WT larval zebrafish using whole‐mount in situ hybridization. ⁴⁰ To further investigate the distribution of the kcnb1 protein in the brain, we performed immunohistochemistry on WT larvae at 6 dpf (Figures 1A–F and S1A,B). Our findings revealed broad expressions of kcnb1 throughout multiple regions of the CNS, such as the diencephalon, midbrain, telencephalon, and hindbrain (Figures 1A–C and S1A) including the spinal cord (Figure 1C), although the protein was difficult to detect in the eyes. Further analysis demonstrated that kcnb1 is expressed in various CNS cell subtypes. Notably, kcnb1 is localized in neurons as evidenced by colocalization with NeuN, a neuronal nuclear marker, and further confirmed by ankyrin G marker, expressed on the axonal initial segment of neurons (Figures 1D and S1B). kcnb1 was also found in oligodendrocytes, marked by Olig2, a transcription factor (Figure 1E), and in microglial cells, identified by CX3CR1, a fractalkine receptor marker (Figure 1F). These results provide the first characterization of kcnb1 expression in distinct CNS cell subtypes in zebrafish.

FIGURE 1.

kcnb1 is expressed in distinct cell subtypes and various regions of the central nervous system in 6 days post fertilization (dpf) wild‐type (WT) zebrafish. (A, B) Horizontal and transversal sections of WT zebrafish expressing cells (4,6‐diamidino‐2‐phenylindole [DAPI]; blue) labeled with anti‐kcnb1 (green), showing a large expression of the protein in the central nervous system (CNS) at 6 dpf. The protein is expressed in the telencephalon, diencephalon, midbrain including the optic tectum, and hindbrain comprising the cerebellum and the spinal cord (A: scale bar = 50 μm, magnification = 10×; B: scale bar = 30 μm, magnification 20×). (C) Horizontal section of a WT zebrafish at 6 dpf showing the presence of kcnb1 along the spinal cord (scale bar = 10 μm, magnification = 63×). (D–F) Horizontal sections of 6‐dpf WT zebrafish expressing cells (DAPI; blue), kcnb1 (green), and specific cell subtype markers, respectively. (D) A neuronal nuclear marker (neuronal nuclear antigen [NeuN]; red, scale bar = 30 μm, magnification = 20×), (E) oligodendrocyte transcription factor 2 (Olig2; red, scale bar = 10 μm, magnification = 63×), and (F) CX3C motif chemokine receptor 1 expressed in microglial cells (CX3CR1; red, scale bar = 10 μm, magnification = 63×). Images demonstrate the colocalization between kcnb1 and the three different cell subtype markers, represented in white and marked by arrows. These results indicate the presence of kcnb1 in neurons, oligodendrocytes, and microglial cells. The colocalization was determined using Z‐stack projection with IMARIS v10.1.0 software (Oxford Instruments). In figures, CNS regions are delimited by dotted lines. D, diencephalon; E, eyes; H, hindbrain; M: midbrain; T, telencephalon. n = 3–4/section.

3.2. kcnb1 KO expression and morphological analyses

Using a kcnb1 KO zebrafish model generated by Shen et al., ⁴⁰ we investigated LOF effects of kcnb1 in the context of DEEs. We first confirmed the genomic sequence of kcnb1 ^−/− fish by Sanger sequencing, identifying a distinct 2‐G bases deletion along with a 14‐bp insertion on chromatograms (∆14 bp; Figure 2A). Reverse transcription qPCR analysis revealed a significant reduction in kcnb1 transcript expression by 44% in kcnb1 ^+/− and by 56% in kcnb1 ^−/− zebrafish larvae compared to WT at 6 dpf (Figure 2B). The survival rates of both kcnb1 ^+/− and kcnb1 ^−/− lines from 0 to 15 dpf were similar to those of WT (Figure 2C and Table S1). Measurements of body length and head surface of kcnb1 KO (kcnb1 ^+/− and kcnb1 ^−/−) showed no significant morphological differences compared to WT at 48 hpf and 6 dpf (Figure 2D–F), corroborated by using the pan‐neuronal marker HuC (Figure S1C). Visualization of neuronal fibers with an acetylated tubulin marker revealed no major neuronal loss or organizational defects due to kcnb1 LOF during early development (Figure 2G), supported by similar Mauthner cell body length observed across conditions at 48 hpf (M‐cell, 3A10 marker; Figure S1D,E). These results indicate that kcnb1 LOF does not impact the normal growth of fish during early development.

FIGURE 2.

Genotypic characterization of a kcnb1 knockout (KO) zebrafish model without brain anatomical defects. (A) Schematic illustration of the generation of a kcnb1 KO zebrafish model obtained by Shen et al. ⁴⁰ The representation of the α‐subunit kcnb1 is adapted from a previous schematic picture. ⁴⁴ Using the CRISPR‐Cas9 (Clustered Regularly Interspaced Short Palindromic Repeats) system, the kcnb1 KO zebrafish model was generated by the insertion of a 14‐bp nucleotide sequence (indicated in red) and of a 2‐bp nucleotide deletion (indicated in black bold type) in the first exon of the gene, resulting in the introduction of a premature stop codon (∆14 bp), and localized between the N‐terminal region and the first transmembrane domain of the protein (targeted sequence: GGAGCTGGACTACTGGGGAG in kcnb1 exon 1; ID zfin: ZDB‐ALT‐170417‐2; indel mutation; line: kcnb1^sq301/sq301). (B) Reverse transcription quantitative polymerase chain reaction analysis of total kcnb1 at 6 days post fertilization (dpf) demonstrating a significant decrease of kcnb1 mRNA in kcnb1 ^+/− and kcnb1 ^−/− compared with wild‐type (WT; kcnb1 ^+/+) larvae (N = 4; n = 30/sample; one‐way analysis of variance [ANOVA] with Bonferroni post hoc test; **p < .01). Data are normalized to actin mRNA expression, and the condition kcnb1 ^+/+ is considered as the reference value (relative fold change = 1). (C) Kaplan–Meier survival curve between 0 and 15 dpf showing that the loss of kcnb1 does not impact the normal growth of fish at early stages of development (see Table S1; N = 3; n = 121–169/genotype; log‐rank test). (D) Images showing that kcnb1 ^+/− and kcnb1 ^−/− zebrafish do not show gross morphological changes at 48 hours post‐fertilization (hpf) and 6 dpf (scale bar = 300 μm). (E, F) Quantification of major morphological aspects at 48 hpf and 6 dpf including measurement of (E) body length and (F) head surface. kcnb1 KO models (kcnb1 ^+/− and kcnb1 ^−/−) do not show any significant change in each parameter at different developmental stages as compared to kcnb1 ^+/+ (N = 4; n = 28–38/genotype; one‐way ANOVA with Bonferroni post hoc test). (G) Whole‐mount images of embryonic (48 hpf) and larvae (6 dpf) zebrafish immunostained with anti‐acetylated tubulin marker to identify global circuits of neuronal fibers (lateral and dorsal view; three‐dimensional reconstruction; magnification = 20× with scale bar at 100 μm; magnification = 40× with scale bar at 20 μm). The loss of kcnb1 does not affect the neuronal brain density at different early stages of development (n = 5–7/genotype/developmental stage). ns, non‐significant.

3.3. Loss of kcnb1 leads to altered behavior phenotype, and light‐ and sound‐induced locomotor impairments

KCNB1 mutations in patients are associated with a range of locomotor disabilities including hyperactivity, myoclonia, ataxia, and hypotonia. ⁹ The first evidence of motor impairments was observed in kcnb1 KO showing a drastic decrease of tail‐coiling activity compared to WT zebrafish from 24 to 36 hpf (Figure S2A,B). Using the TEER test at 48 hpf (Figure 3A), kcnb1 ^+/− and kcnb1 ^−/− zebrafish exhibited rapid circular swimming, in contrast to the straight‐line swimming observed in WT condition (Figure 3B, Video S1a,b), significantly increased in terms of total distance swam, velocity, and time spent in motion (Figure 3C–E). Notably, within the same condition, we observed a variability in the swimming behavior of embryos classified into two groups: those with a severe phenotype (completing a minimum of two swim circles) and those with a mild phenotype (grouping other trajectories; Table S2). We found that 30% of kcnb1 ^+/− embryos displayed a severe phenotype compared to 52% of kcnb1 ^−/− zebrafish, with an increase of studied parameters (Table S2, Figure S2C–F). To validate the locomotor phenotype observed at this stage, kcnb1 ^+/− and kcnb1 ^−/− eggs were injected with hK _v 2.1‐WT. Post‐mechanical stimulus analysis revealed a rescue of the swimming phenotype, characterized by straight‐line or single‐circle trajectories comparable to WT embryos, and associated with a significant reduction in swimming distance (Figure 3B,F). Notably, none of the kcnb1 ^+/− or kcnb1 ^−/− embryos exhibited the severe locomotor phenotype described in Table S2 following the rescue. At later developmental stages, although all three conditions showed low locomotor activity in response to light changes at 3 and 4 dpf, this activity significantly increased in all conditions at later stages (Figure 3G). However, starting at 5 dpf, kcnb1 ^−/− larvae displayed pronounced locomotor hyperactivity compared to WT and kcnb1 ^+/− zebrafish (Figure 3G). This finding is aligned with results from Figure 3H,I, showing significant locomotor hyperactivity in response to light/dark transitions. In a second protocol, both kcnb1 ^+/− and kcnb1 ^−/− KO exhibited a significant decrease in locomotor activity after each audio stimulus compared to the WT condition (Figure 3J). These results suggest that the locomotor phenotype may be due to dysregulation of electrical neuronal activity affecting different sensorimotor pathways.

3.4. kcnb1 ^−/− zebrafish exhibit increased sensitivity to PTZ‐induced seizures and elevated expression of epileptogenesis‐related genes

Previous studies have shown that exposure of zebrafish larvae to PTZ increases seizure‐like behavior in a concentration‐dependent manner. ³⁰ We recorded the baseline locomotor activity of 6 dpf larvae for 30 min, followed by a 30‐min exposure to 5 mmol·L⁻¹ PTZ (Figure 4A). Low baseline locomotor activity of kcnb1 KO was similar to that of WT larvae (Figure 4B,C). However, after PTZ exposure, kcnb1 KO (kcnb1 ^+/− and kcnb1 ^−/−) exhibited significantly higher swimming activity, characterized by fast circles (Figure 4B,C and Video S2). To explore potential molecular disruptions, we measured bdnf (brain‐derived neurotrophic factor) mRNA expression, an epileptogenesis‐related gene, before and after PTZ treatment (Figure 4A,D). bdnf expression was similar between conditions during baseline activity, although significantly increased after PTZ treatment in the kcnb1 ^−/− condition as compared to kcnb1 ^+/− and WT zebrafish (Figure 4D). In addition, we assessed c‐Fos expression, an early gene marker of epileptic seizures, in the telencephalon of 6‐dpf larvae (Figures 4A,E–G and S3). The number of c‐Fos‐positive neurons was similar across all conditions, both before and after PTZ treatment (Figure 4F). We analyzed the distribution of activated neurons according to the fluorescence intensity value of c‐Fos, reflecting the level of neuronal activation divided into low (0%–25%), moderately low (25%–50%), moderately high (50%–75%), and high (75%–100%) neuronal activation (Figures 4G and S3B). kcnb1 ^−/− fish showed a shift toward higher levels of neuronal activation, with a significant increase in moderately low and moderately high activation levels after PTZ treatment (Figure 4G and S3B). These results suggest that the kcnb1 LOF model exhibits neurogenesis impairments in the developing brain of zebrafish.

3.5. kcnb1 KO zebrafish model shows spontaneous and provoked "epileptic‐like" seizures associated with disrupted γ‐aminobutyric acid regulation

Abnormal electrographic activity has been observed in zebrafish models with chemically provoked epileptic seizures. ³⁰ , ³⁵ , ³⁸ We recorded LFPs from the optic tectum of 6‐dpf larvae during a 30‐min “basal activity” phase, followed by a 40mmol·L⁻¹ PTZ treatment, recorded for an additional 30 min (Figure 5A,B). Prior to PTZ treatment, kcnb1 ^−/− larvae showed significantly increased spontaneous neuronal activity, reflected by the number of spikes, compared to WT (Figure 5Ca–D). In contrast, kcnb1 ^+/− larvae did not exhibit spontaneous seizures, displaying a similar profile to untreated WT zebrafish (Figure 5Ca–D). In response to 40mmol·L⁻¹ PTZ exposure, WT larvae exhibited a significant increase in number of spikes compared to the nontreated condition (Figure 5Ca–D). The kcnb1 ^+/− model showed a similar response profile, although the duration of seizure events was significantly longer than in PTZ‐treated WT larvae (Figure 5Ca–E). Interestingly, kcnb1 ^−/− larvae had a significantly higher number of PTZ‐induced events, although event duration was comparable to WT‐treated zebrafish (Figure 5Ca–E). This kcnb1 ^−/− model also displayed various "epileptic‐like" signals seen in KCNB1‐related DEE patients, including "ictal‐like" and "interictal‐like" activity (Figure 5Cb,Cc), polyspike discharges (Figure 5Cd), and large‐amplitude spikes (Figure 5Ce). To further confirm dysregulated neuronal activity in the kcnb1 LOF model, we measured γ‐aminobutyric acid (GABA) and glutamate concentrations in the heads of 6‐dpf zebrafish following the same protocol and exposed to 5 mmol·L⁻¹ PTZ (Figure 5A). During the pretreatment period, GABA levels were similar in kcnb1 ^+/− and WT larvae, whereas kcnb1 ^−/− larvae showed a significant increase (Figure 5F). Although PTZ treatment significantly increased GABA concentration in WT zebrafish, it did not alter GABA levels in the kcnb1 ^+/− model, which remained low, or in the kcnb1 ^−/− model, showing similar high concentration of GABA as WT‐treated larvae (Figure 5F). Glutamate levels were similar between all genotypes before PTZ exposure and remained unchanged posttreatment (Figure 5G). These results suggest that the kcnb1 ^−/− model exhibits spontaneous and chemically induced epileptiform‐like electrographic activity, along with disrupted GABA regulation.

4. DISCUSSION

In this study, kcnb1 KO in zebrafish results in early onset phenotypes mimicking key features of KCNB1‐related DEE. This LOF impacts neuronal functions, in particular inhibitory pathways in developmental brains.

4.1. kcnb1 expression in the CNS

The study of kcnb1 expression revealed the presence of the protein across various regions of the zebrafish CNS, including the diencephalon, midbrain, telencephalon, and hindbrain, consistent with previously reported in situ hybridization of kcnb1 by Shen et al. ⁴⁰ This broad expression supports previous findings that kcnb1 is crucial for maintaining neuronal excitability, starting from 19 hpf. ³⁹ Notably, the presence of kcnb1 in multiple cell subtypes suggests its diverse functional roles, ranging from neuronal signaling to potential involvement in glial cell function and neuroinflammation in zebrafish.

4.2. kcnb1 KO and developmental consequences

Despite a significant reduction in kcnb1 transcript levels in both heterozygous and homozygous KO conditions, we observed no major morphological, brain anatomical abnormalities or impaired survival rates in kcnb1 ^−/− zebrafish during the early development. This finding is consistent with clinical observations in KCNB1‐related DEE patients, where normal brain morphology is reported in most cases despite severe neurological symptoms. ⁹ This suggests that kcnb1 LOF does not impact brain structural development, although it may influence more neuronal functional aspects.

4.3. Altered behavioral phenotypes and sensorimotor dysregulation

Behavioral assays revealed that kcnb1 ^−/− zebrafish exhibit significant locomotor impairments, including hyperactivity, altered swimming patterns, and exaggerated responses to sensory stimuli such as light and sound. We showed that kcnb1 ^+/− and kcnb1 ^−/− embryos had reduced tail‐coiling activity from 25 hpf to 31 hpf. This early stereotyped behavior is supported by synchronized spinal locomotor circuits. ⁴⁵ Previous reports on models of epilepsy have revealed a corkscrew‐like trajectory swimming characteristic of epileptogenic‐like activity. ²⁶ , ³³ We found a similar circular pattern of swim trajectories in kcnb1 KO models at 48 hpf, which was rescued by the presence of the hK _v 2.1‐WT DNA. However, larvae revealed a huge variability within the same genotype. These results might reflect the variable spectrum of behavioral and cognitive impairments observed in patients. ¹² Furthermore, kcnb1 ^−/− KO exhibited increased response of locomotor activity to a light–dark stimulus as compared to kcnb1 ^+/− and WT lines but presented decreased locomotor activity after a succession of audio stimuli. This last result is in correlation with the data obtained by Jedrychowska et al. ³⁹ These phenotypes closely mimic the motor dysfunctions observed in patients with KCNB1 mutations, such as ataxia and hyperactivity, and suggest a dysregulation of sensorimotor pathways.

4.4. Seizure susceptibility and electrophysiological abnormalities

The seizure‐like behavior in kcnb1 ^+/− and kcnb1 ^−/− models is supported by their increased locomotor activity induced by 5 mmol·L⁻¹ PTZ, corresponding to stage I of the "seizure‐like behavior score" as previously described. ³⁰ In most cases, larvae then exhibited fast circular trajectories characterized by rapid "whirlpool‐like" movement and a brief pause (stage II). However, none of the kcnb1 KO larvae displayed stage III, defined by loss of posture and immobilization, potentially reflecting tonic/clonic‐like seizure. These results are concomitant with the “seizure‐like behavior score” described by Baraban et al. ³⁰ under similar PTZ treatment conditions. Our electrophysiological recordings further demonstrated that kcnb1 ^−/− zebrafish exhibit spontaneous and chemically induced epileptiform seizure‐like activity. The absence of spontaneous seizures in the kcnb1 ^+/− line could reflect the existence of functional compensatory pathways masking the phenotypic features due to reduced levels of kcnb1 mRNA but similar protein expression to WT larvae. Epileptiform seizure‐like activity was confirmed by c‐Fos acute neuronal hyperactivation in the telencephalon of kcnb1 ^−/− zebrafish. Moreover, the quantification of bdnf mRNA in larvae, a neurotrophin crucial for brain development and synaptic plasticity, revealed a significant upregulation of its expression level within the kcnb1 ^−/− line. This finding indicates deficits in neurogenesis within the developing brain of kcnb1 ^−/− zebrafish. Similar results were obtained by identifying c‐Fos and bdnf as genes associated with seizures in zebrafish, with increased expression levels observed after exposure to 20 mmol·L⁻¹ PTZ. ⁴⁶ These findings are particularly relevant to understanding the pathophysiology of epilepsy in DEEs, as zebrafish exhibit spontaneous and chemically provoked seizures.

4.5. Neurotransmitter dysregulation in kcnb1 ^−/− zebrafish

Our study also highlights the dysregulation of γ‐aminobutyric acid (GABA) in kcnb1 ^−/− zebrafish, with significantly elevated GABA levels observed in both baseline and post‐PTZ conditions. This disrupted GABAergic signaling likely contributes to the observed seizure phenotypes, as GABA is a key inhibitory neurotransmitter involved in maintaining the balance of excitatory and inhibitory signals in the CNS. The lack of significant changes in glutamate levels suggests that kcnb1 LOF primarily affects inhibitory pathways, leading to an imbalance that favors neuronal hyperexcitability. The high‐throughput nature of zebrafish drug screening allows for the rapid evaluation of potential antiepileptic drugs, including those that modulate GABAergic transmission. This approach holds particular clinical relevance for KCNB1‐related disorders, where impaired network excitability contributes to epilepsy. In this context, further studies targeting the modulation of GABAergic signaling are currently being investigated.

4.6. Study observations

Although the kcnb1 KO zebrafish model reproduces certain pathophysiological features observed in patients, the genetic modification in this vertebrate model differs from the genetic conditions associated with KCNB1‐related DEE. ⁹ , ⁴⁰ The mutation induced by CRISPR/Cas9 results in a premature termination codon located in exon 1, potentially triggering nonsense‐mediated decay. ⁴⁰ This deletion mutant zebrafish model contrasts with the majority of pathogenic variants in KCNB1‐related DEE, which involve a single amino acid substitution in exon 2. ⁹ Therefore, further cellular and animal models will be needed in the future to better understand disease mechanisms of KCNB1‐related DEE and to develop therapeutic strategies for this rare neurodevelopmental disorder. Recent clinical studies provide valuable context for understanding the human phenotype spectrum of KCNB1 mutations, detailing the diversity of comorbidities and seizures, including myoclonic–atonic seizures, infantile spasms, generalized tonic–clonic seizures, and focal seizures. ¹² , ¹⁴ , ¹⁵ By contrast, the zebrafish model predominantly exhibits early stage hyperlocomotor activity and seizure‐like activity, which simplifies the phenotype but may not capture the full complexity of KCNB1‐related DEEs. This underscores the importance of interpreting zebrafish data within the context of human pathophysiology while leveraging the model's strengths for early stage discovery. These differences may be due to differences between species in brain development, where zebrafish exhibit a simpler brain architecture, lacking a neocortex and possessing less intricate neuronal networks. Additionally, our study primarily focused on defining phenotypic features in early developmental stages. It would be interesting to investigate the neuronal and epileptic status of kcnb1 heterozygous and homozygous zebrafish at juvenile and adult stages to identify complex seizure‐like behavioral phenotypes. This extended examination could provide a more comprehensive understanding of the long‐term impact and manifestations associated with kcnb1 LOF in the zebrafish model.

5. CONCLUSIONS

Our investigation suggests that the kcnb1 LOF zebrafish model provides a valuable model to reproduce early behavioral disturbances, increased susceptibility to epileptic seizures, and neurotransmitter dysregulation. Notably, this model could be used for discovering new therapeutic compounds that may improve the long‐term prognosis of individuals with KCNB1‐related DEE.

AUTHOR CONTRIBUTIONS

Lauralee Robichon designed and conceptualized the study, performed the experiments, analyzed and interpreted the data, and wrote the manuscript. Claire Bar designed some parts of the study, performed some behavioral assays, and wrote the manuscript. Anca Marian performed some behavioral assays and was responsible for technical support. Lisa Lehmann performed some locomotor activity experiments and analyzed some data. Solène Renault was responsible for technical support. Edor Kabashi supervised the work, revised the paper, and approved the manuscript. Sorana Ciura supervised the work, revised the paper, and approved the manuscript. Rima Nabbout supervised the work, revised the paper, and approved the manuscript. No undisclosed groups or persons have had a primary role in the study and/or in manuscript preparation. All coauthors have been substantially involved in the study and/or the preparation of the manuscript. All coauthors revised and approved the submitted version of the paper and accept responsibility for its content.

CONFLICT OF INTEREST STATEMENT

None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Supporting information

Figure S1.

Figure S2.

Figure S3.

Table S1.

Table S2.

Table S3.

Video S1.

Video S2.

Text S1.

Robichon L, Bar C, Marian A, Lehmann L, Renault S, Kabashi E, et al. kcnb1 loss of function in zebrafish causes neurodevelopmental and epileptic disorders associated with γ‐aminobutyric acid dysregulation. Epilepsia. 2025;66:3048–3063. 10.1111/epi.18407

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.