Progress and potential of brain organoids in epilepsy research

Abstract

Epilepsies are disorders of the brain characterised by an imbalance in electrical activity, linked to a disruption in the excitation and inhibition of neurons. Progress in the epilepsy research field has been hindered by the lack of an appropriate model, with traditionally used 2D primary cell culture assays and animal models having a number of limitations which inhibit their ability to recapitulate the developing brain and the mechanisms behind epileptogenesis. As a result, the mechanisms behind the pathogenesis of epilepsy are largely unknown. Brain organoids are 3D aggregates of neural tissue formed in vitro and have been shown to recapitulate the gene expression patterns of the brain during development, and can successfully model a range of epilepsies and drug responses. They thus present themselves as a novel tool to advance studies into epileptogenesis. In this review, we discuss the formation of brain organoids, their recent application in studying genetic epilepsies, hyperexcitability dynamics and oxygen glucose deprivation as a hyperexcitability agent, their use as an epilepsy drug testing and development platform, as well as the limitations of their use in epilepsy research and how these can be mitigated.

Background

Epilepsies are a diverse family of disorders of the brain characterised by the susceptibility to generate epileptic seizures [1]. These disorders are the most common cause of neurological disease and affect 50 million people worldwide, which corresponds to 0.5% of the global burden of disease [2, 3]. The prevalence of epilepsy hinges on the economic status of a country, with the incidence being higher in both low and lower-middle-income countries compared to higher income countries [2]. However, within these economic settings, epilepsy affects people irrespective of their gender, age, ethnic background or geographic location [3]. The causes of epilepsy can range from genetic and metabolic, to infectious, structural and immune, or may even be undetectable, and is accompanied by cognitive, neurobiological, psychological and social consequences [1, 3]. The term epileptogenesis has been used to describe the development of the state of epilepsy. This occurs when there is a disruption in the excitation and inhibition of neurons within the brain, which creates a conducive environment for seizures [1, 4]. Mechanisms that promote the increase in excitation and the disruption in the inhibition of this excitation, allow the brain to develop an imbalance in electrical activity, resulting in large groups of neurons becoming hyperexcitable and abnormally discharging [4]. This alteration in electrical activity of the brain creates an environment that increases the occurrence and likelihood of future seizures [1].

The increased likelihood of seizures presents a number of concerns related to their impact on other bodily functions, in addition to societal implications. These range from adverse effects on autonomic, motor and sensory functions as well as consciousness, memory, cognition, behaviour, and emotional state [1]. Learning disabilities and academic weakness are also prevalent in children with epilepsy [5]. Generally, individuals with epilepsy also face social consequences, including stigmatization and discrimination, and these consequences are exacerbated when individuals are left untreated [1, 3]. In addition, approximately 50% of epilepsy sufferers have accompanying psychiatric or physical conditions, with the most common psychiatric comorbidities being depression and anxiety, present in 23% and 20% of cases, respectively [3].

These statistics demonstrate the great need for continued research in the field of epilepsy, especially in order to understand the underlying mechanisms behind epileptogenesis, to discover therapeutic drug targets, and to test novel drugs for the treatment and management of this condition. Here, we outline the current models used to study epilepsy and their limitations as well as reviewing the promise of using brain organoids as a model in assessing neuronal hyperexcitability in epilepsy, whilst also addressing their limitations.

Current models for studying epilepsy and their limitations

There are several models that are currently used to study epilepsy with varying degrees of success (Fig. 1). Ideally, one would use human brain tissue itself to study epilepsy, which can be retrieved during surgery or post mortem [6]. However, due to a lack of access to surgical tissue as well as significant deterioration that occurs when using post-mortem samples, much research has explored the use of animal models of epilepsy [7]. Acute and organotypic animal slice assays, in which slices of brain tissue are obtained from any species with a complex brain, have been used in a range of short-term and long-term experiments, with the longer term experiments becoming more accessible due to the development of various methods to increase the lifespan of the resected tissue [8, 9]. These improved methods have enabled the maintenance of organotypic slices of neonatal rat tissue in vitro for several weeks. With the added benefit of the retained cytoarchitecture of the in vivo brain in the tissue sections, they are a favourable model for epilepsy research [10, 11].

Fig. 1.

Current models for studying epilepsy with their applications and limitations

Animal models have also been used in a broad range of studies which have helped to uncover the potential mechanisms behind epileptogenesis and seizure development [4], the deficits in cognition and behaviour associated with seizures in early life [12, 13], as well as the effectiveness of different convulsant and anti-convulsant drugs for the treatment of epilepsy [14, 15]. Many of these studies make use of epilepsy mouse models, in which genetic mutations that are associated with epilepsy are introduced into the mouse genome before birth [14, 15]. This has helped researchers to gain insights into the role of different genetic mutations in epileptogenesis and how this can be managed using different drugs.

Alternative models that are used in smaller studies are primary cell culture assays using mouse or human cells which can be cultured with relative ease, reduce the costs and complexity associated with the use of the models mentioned above, and produce data that can be compared and contrasted with both animal models and tissue slice assays [10, 16]. By providing neuronal and glial cell cultures with the precise culture conditions that mirror the in vivo microenvironment, study of the central nervous system (CNS), action potentials and synapses to gain insight into the pathophysiology of epilepsy and the mechanism of action of several antiepileptic drugs has been possible [16].

The use of the above models has enabled major advancements in the epilepsy research field, however, they have limitations. The main challenges that arise are the inaccessibility of healthy and diseased brain tissue as well as the ethics involved with working with it, which result in lower throughput capabilities of these studies [17]. The severing of projection neurons and ischemia of the tissue upon removal also limits the use of resected tissue [10]. In addition to this, generating more relevant adult models also presents challenges with maintaining long-term culture conditions and sterility, both of which are time consuming, laborious and require skilled personnel [18], not to mention the difficulties with genetically manipulating the tissue due to its complexity [19, 20].

Although the use of animal models, including animal primary cell culture and brain tissue slices, have facilitated developments in our understanding of epilepsy, interspecies differences remain [10], with the species-specific features of the brain presenting challenges for the extrapolation of the data to predict outcomes in a human model [19, 20]. Animal tissue slices in particular are often also resected from neonatal rats, which means that they don’t accurately represent adult tissue [18], whereas in the case of animal primary cell culture assays, they have their own difficulties linked to harvesting and maintaining neurons in vitro and the lack of proliferation of neurons in these cultures [21].

In the past, human and animal brain tissue models have been heavily relied on for neurodevelopment and neurotoxicity testing. However, due to the ethical concerns, resource availability and cost associated with the use of these models, there is a need for an improved model that more closely represents the in vivo brain and can be used in studies of neurodevelopment and the development of neurological disorders, such as epilepsy.

Formation of brain organoids

One specific cell type that is increasing in popularity for use in primary cell cultures are human induced-pluripotent stem cells (hiPSCs). These cells have several advantages which help them to overcome challenges that arise when working with neural tissue. hiPSCs provide the means to separately differentiate neuronal cell types, which have been shown to recapitulate components of the cortex and neural progenitor cells [22], and they can be cultured with the underlying genetic variants associated with a specific disorder for disease models [6, 21]. Although culture of these cells is advantageous, they lack the 3D architecture of the brain, rendering them unable to represent a realistic in vivo CNS microenvironment with surrounding supporting cell types [18]. Recently however, the advent of 3D models of the human brain, so-called brain organoids, have presented themselves as a more structurally and morphologically relevant model in which to study epilepsy [22–27].

Brain organoids are 3D aggregates of neural tissue, formed from hiPSCs or human embryonic stem cells (hESCs), which are generated in vitro by differentiation and self-organisation to initially form neuroectoderm, then differentiate into expanded neuroepithelium and finally into cerebral tissue (Fig. 2) [19, 24, 28]. To generate brain organoids, stem cells are provided with neural induction medium which directs them to differentiate into neuroectoderm within embryoid bodies (EBs) [25]. Subsequently, the EBs are placed in differentiation medium in which they form neural rosette structures that grow and expand into neuroepithelium and cerebral tissue [22, 25]. Spinning bioreactors enable long-term culture of cerebral tissue through the increased diffusion of nutrients into the core of the organoid, allowing the organoids to better represent later stages of neurodevelopment in vitro [22, 29].

Fig. 2.

Formation of brain organoids (A) Human induced pluripotent stem cells (hiPSCs) are derived from skin fibroblasts taken from a skin biopsy. (B) Human embryonic stem cells (hESCs) are taken from the inner cell mass of the blastocyst. Both these cell types are used to generate (C) embryoid bodies (EBs) that can be differentiated into a range of organoid types that represent various areas of the brain. Cerebral organoids represent the cerebrum of the brain, oligocortical spheroids have a higher representation of oligodendrocytes, while forebrain organoids represent the forebrain region of the brain. Single rosette spheroids specifically contain only a single neural rosette in the centre as opposed to multiple in classic brain organoid protocols. Lastly, asteroids contain a higher representation of astrocytes

There are a range of subtypes of brain organoids and they are classified based on their composition, the region of the brain they represent, and the way in which they are generated. Some of these include oligocortical spheroids, asteroids, and forebrain organoids, as well as single rosette spheroids [20, 30–37]. The physiological relevance of these models has been widely tested but one such example used RNA-seq to show that organoid cortical cells have a gene expression pattern that closely mirrors the brain, demonstrating the promise of brain organoids for studying disease phenotypes [19, 38]. Because of their promise as an in vitro model, brain organoids have been used in epilepsy research over the last few years. However, given that the generation of brain organoids is a relatively new field, only a few studies have been conducted on the neuronal hyperexcitability implicated in epilepsy specifically.

Investigating hyperexcitability in brain organoids

There are a range of studies that explore the use of electrophysiological methods to study the hyperexcitability of brain organoid models in an attempt to uncover the mechanisms behind epileptogenesis (Table 1). The most commonly used method for measuring this hyperexcitability is whole-cell patch-clamp electrophysiology. This technique allows the study of the electrical properties of neurons and can be used to measure either single or multiple ion channel functions of tissues [39]. In 2023, Landry et al. performed whole-cell patch-clamp as well as a range of other electrophysiological recordings on intact brain organoids [40]. This allowed them to determine both the intrinsic and active electrophysiological properties of the cells within the organoids. The ability of this method to routinely characterise single neurons within their intact networks and circuits has significant implications for epilepsy research and other disease studies linked to dysfunction in synapse formation, excitability and circuit activity of neurons [41]. Not only is this a very effective protocol, but it also avoids the possible confounding effects associated with making these measurements in 2D cultures such as culture physiology, species-specific responses and behavioural assays, as well as the damage due to slicing or dissociating cultures for these measurements, altogether rendering this as a very promising model [42].

Table 1.

Summary of papers that model epilepsy in organoids and their main findings

Organoid model	Disease modelled	Electrophysiology measurement approach	Main findings	Reference
Brain organoid	N/A	Whole-cell patch-clamp recordings	Whole-cell patch-clamp was able to characterise neurons from intact human brain organoids, as opposed to acute slices of dissociated cultures, based on their morphological and electrophysiological profile.	Landry et al. [40]
Brain organoid	CDKL5 deficiency disorder (CDD) and Rett Syndrome (RTT)	Whole-cell patch-clamp recordings	CDD organoids had increased driving force for action potential firing which led to higher action potentials. CDD organoids also had a faster depolarisation and repolarisation rate. The researchers uncovered a critical role of neural hyperexcitability and ion channel dysfunction in early brain development in both CDD and RTT disorders.	Wu et al. [43]
Oligocortical spheroid	WWOX-related epileptic encephalopathies (WOREE) syndrome	Whole-cell patch-clamp and local field potential recordings	Local field potential recordings were higher for the disease model showing enhanced neocortical excitability and an increase in low-frequency activity.	Repudi et al. [37]
Cerebral organoid	Developmental and epileptic encephalopathies (DEE)	Local field potential recordings	Disease organoids had a higher power and decreased oscillatory power over time, as well as hyperexcitability and increased activity in response to convulsant drug administration.	Steinberg et al. [26]
Cerebral organoid	Hypoxic-ischemic encephalopathy (HIE) induced through oxygen-glucose deprivation (OGD)	Whole-cell patch-clamp recordings	The organoids were able to show increases in power spectrum changes in response to various chemoconvulsants. Increases such as these are helpful in identifying seizure onset zone due to their link to higher propensity for hyperexcitable neural activity.	Saleem et al. [61]
Cerebral organoid	Hypoxic-ischemic encephalopathy (HIE) induced through oxygen-glucose deprivation (OGD)	Acute slice electrophysiology	OGD was effective in inducing a hyperexcitable state in the immature tissue, which was demonstrated by the increase in power spectrum compared to baseline in the organoids.	Santos et al. [62]
Cerebral organoid	N/A	Multielectrode array	The cerebral organoids had sustained neural activity and a concentration-dependent seizure-like waveform in response to drugs, with the response varying between drugs administered.	Yokoi et al. [66]
Cerebral cortex-ganglionic eminence (Cx+GE) fusion organoid	Rett Syndrome	Calcium sensory imaging and local field potential recordings	Fusion organoids had functionally integrated excitatory and inhibitory neurons. Local field potential recordings uncovered simultaneous sustained oscillations at multiple frequencies, demonstrating the formation of mature neural networks.	Samarasinghe et al. [67]
Single-rosette spheroid cerebral organoid (SOSR-CO)	Protocadherin-19 clustering epilepsy (PCE)	Whole-cell patch-clamp recordings	SOSR-COs showed increased action potential firing and burst activity over time. They also saw an increase in electrophysiological network maturation over time, beginning at 3 months-old.	Tidball et al. [36]

Whole-cell patch-clamp recordings have also been used in other disease studies which are associated with epilepsy and varying mechanisms of epileptogenesis (Fig. 3). Wu et al. (2022) generated hiPSC-derived cortical brain organoids from two patients with cyclin-dependent-kinase-like 5 (CDKL5) deficiency disorder (CDD) and used whole-cell patch-clamp techniques to investigate the synaptic transmission and action potential firing properties of the neural cells in these organoids compared to controls [43]. CDD is a neurodevelopmental disorder that arises when there are loss-of-function mutations in CDKL5 [44–46]. Patients with CDD present with neurodevelopmental delay, early-onset seizures, and autistic behaviours due to neuron hyperexcitability and ion channel dysfunction [47, 48]. Prior to this study, the mechanism behind the excessive excitability seen in this disorder was largely unknown. They noted that the hyperexcitability phenotype of CDD can be recapitulated in a 3D brain organoid model, allowing them to successfully conclude that ion channel dysfunction and neural hyperexcitability seen in both CDD and another neurodevelopmental disorder called Rett Syndrome play a critical role in disease pathogenesis. They also discovered that there was a convergence between the two neurological diseases in regard to their electrophysiological signatures which suggests that the mechanism behind seizure generation and epileptogenesis between these two disorders overlaps.

Fig. 3.

Whole-cell patch-clamp (A) Representation of whole-cell patch-clamp used to measure the activity of ion channels in a single neuron. A recording micropipette is placed in contact with the neuronal cell membrane which forms a seal with the membrane. This allows the measurement of tiny current-voltage changes in the neuron and the intracellular recording of ion channel activities across whole cell membranes. (B) Graphs generated from whole-cell patch-clamp show current change in the neuron over time. (C) In epilepsy, neurons become hyperexcitable and depolarise and repolarise more frequently, resulting in higher current output and depolarisation to the lesser extent

Studying genetic epilepsies using brain organoids

Brain organoids used in combination with CRISPR-Cas9 has allowed for the robust study of genetic epilepsies through the generation of disease organoids. In 2021, Repudi et al., generated oligocortical spheroids from hESCs that were genetically engineered to create a neuronal WW domain–containing oxidoreductase (WWOX) knockout model, which is associated with WWOX-related epileptic encephalopathies (WOREE) syndrome [37]. This is a rare developmental disorder that is characterised by both global developmental delay and the presence of drug-resistant epilepsy, and arises when there are germline mutations in the WWOX tumour suppressor gene [49–51]. A conditional deletion of WWOX in neural progenitors, neural stem cells or matured neurons can result in epileptic seizures and growth retardation. These mutations can also give rise to spinocerebellar ataxia type 12 (SCAR12 syndrome) [52, 53]. For the completion of this study, whole-cell patch-clamp and local field potential recordings were done on WWOX knockout oligocortical spheroid slices. The researchers noticed that complete deletion of WWOX was associated with hyperexcitability and hypomyelination, which confirmed the role of the neuronal functions of WWOX in epilepsy and WOREE syndrome. These spheroids were able to successfully model epileptiform activity and confirm the data that this group saw in their mouse studies, which links WWOX loss in the brain and the pathogenesis of epilepsy.

A study conducted by Steinberg et al. (2021) also looked at WOREE syndrome, where they used it as a prototype for developmental and epileptic encephalopathies (DEE) [26]. These are a group of disorders that present in the form of seizures as well as impaired brain development and cognitive functions [54–57]. A strong link has been established between autosomal recessive mutations in the WWOX gene and DEE. For this study, the researchers made use of CRISPR to introduce a germline WOXX gene knockout in hESCs, which they then used to generate cerebral organoids. They used local field potential recordings to measure the epileptiform activity present, and observed hyperexcitability in both their WWOX knockout cerebral organoids as well as patient-derived cerebral organoids. Additionally, they were able to demonstrate imbalances between gamma-aminobutyric acid (GABA)-ergic and glutamatergic neurons, defected DNA damage responses, chronic activation of Wnt, enhanced astrogenesis and cortical dysplasia, which all linked to the phenotype of these genetic epileptic encephalopathies, once again revealing the potential of brain organoids to model epileptiform activity.

The effect of oxygen-glucose deprivation on brain organoids

Despite the recent success of brain organoids in epilepsy research, a number of epileptogenetic traits are still poorly understood. One such example is the mechanism behind the development of neonatal seizures. In an attempt to elucidate the answer to this question, scientists have made use of various stimulants to induce hyperexcitability. One of the most effective stimulants is oxygen-glucose deprivation (OGD). Neuronal hyperexcitability has been shown to be associated with OGD and can lead to hypoxic-ischemic encephalopathy (HIE) [58]. HIE is brain damage that occurs in response to OGD, and has a strong link with the high incidence of seizures that occur in the neonatal period [59, 60]. In 2023, a research group set out to use OGD in 3–7-month-old cerebral organoids as an excitability agent, and compared this to a range of other chemoconvulsants [61]. To measure the effect of this, they did whole-cell patch-clamp on acute organoid slices. Their findings confirmed that OGD is indeed the most effective inducer of hyperexcitability in cerebral organoids and this is in agreement with WWOX knockout genetic models of epilepsy in other studies [26]. Another group also made use of OGD in their own human cerebral organoid models [62]. They also demonstrated the stimulating effect of OGD on the hyperexcitability of their organoids by showing an increase in the power spectrum. Additionally, they tested two other convulsants to see which would have a similar effect to OGD. They saw that bumetanide and cannabidiol are both effective at inducing a similar hyperexcitable phenotype in the cerebral organoids. Not only is OGD effective for studying the mechanism of epileptiform activity in cerebral organoid models, but it can also be used to demonstrate the ability of organoids to model seizures in response to a stimulus and evaluate the effect of a range of different drugs compared to a well-known stimulant. This shows great advancement in the field of organoid and epilepsy research, but it is important to note that further characterisation of this model is important to increase its reproducibility and reliability.

Drug testing in brain organoids

Drug testing and improvement for epilepsy treatment is a very important field and brain organoids have shown great success for use in these studies [59, 63–65]. Because the mechanisms behind epileptogenesis are poorly understood, it is challenging to produce effective drugs. In order to increase the efficiency of drug screening, Yokoi et al. (2021) made use of a planner multielectrode array (MEA) coupled to cerebral organoids to evaluate their response to convulsants and specific drugs with different mechanisms of action [66]. This method was proven useful in measuring the seizure liability of drugs and the changes in activity of the organoids after administration of the antiepileptic drugs (AEDs). They were then able to determine which drug had the most successful mechanism of action in normalising the hyperexcitable state in the organoids. This MEA method not only has implications for drug testing but also implications for increasing the throughput of all organoid experiments.

Alternative brain organoid systems in epilepsy research

While cerebral organoids have been the preferred organoid system used in epilepsy studies, they still lack many of the cell types present in vivo, potentially reducing their relevance in a clinical setting. In order to address this, Samarasinghe et al. (2021) generated cerebral cortex-ganglionic eminence (Cx + GE) fusion organoids [67]. In this protocol, select organoids were administered Smoothened Agonist, which stimulated the ventralising activity of the Sonic Hedgehog pathway (Shh), to induce the formation of GE progenitors and migratory interneurons and generate GE organoids. On the contrary, in the absence of activation of the Shh pathway, the organoids developed cortical characteristics to form Cx organoids. Fusion of the two organoid types integrated both inhibitory and excitatory neurons in a functional manner, allowing them to identify epileptiform and neural oscillation changes. Calcium sensory imaging and extracellular recordings were used to measure the activity of the circuits in the Cx-GE organoids, showing that you can fuse organoids and use them to investigate the workings behind network-level functions in the brain.

A concern raised in organoid studies is the spontaneous nature by which they form, resulting in organoids with multiple neural rosette structures, distinctly different from what would occur during in vivo brain development [22, 68, 69]. To address this, Tidball et al. (2022) generated single-rosette brain organoids from hiPSCs [36], and in the process they were successful in generating single-rosette spheroid cerebral organoids (SOSR-COs) that were self-organising and of a reproducible size. These spheroids demonstrated consistent size as well as the cellular diversity and inside-out cortical lamination that occurs during normal development. Additionally, they were able to use these SOSR-COs to successfully model a genetic neurodevelopmental epileptic disorder, protocadherin-19 clustering epilepsy (PCE). They made use of whole-cell patch-clamp techniques on dissociated neurons as well as MEA plates to measure the firing rate of action potentials. They saw that the SOSR-COs were able to successfully model the developmental and structural brain abnormalities associated with PCE, indicating that this system has great promise as a model of neurological diseases, including epilepsies.

Overcoming the limitations of brain organoids

While brain organoids present as an exciting advancement for epilepsy research, they are not without their limitations in this field. A principal concern of brain organoids is their limited maturation. When iPSCs are reprogrammed, they revert to a foetal-like state [70], consequently organoids produced from these cells represent a neonatal as opposed to an adult brain [70]. There are a number of solutions to this problem, and these include integrating supporting cells, such as endothelial cells around the organoid [27], artificially aging the stem cells by overexpressing factors such as progerin, slice culturing the organoids, growing the tissue on microfluidic chip platforms or lengthening the culture period of the organoids [70]. Lengthening of the culture period seems to be the most obvious solution to this problem but due to the lack of vascularization of the organoids, this is challenging [71, 72]. As the culture period lengthens and the organoid grows, diffusion of nutrients is no longer able to reach the inner cells. As a result, the core of the organoid lacks the perfusion of growth medium and vital gases for survival, and the core undergoes necrosis, limiting the culture period of the organoid [73]. Some solutions to this problem include making recordings from the outer edge of the organoid where the cells have not undergone necrosis or introducing a vascular scaffold surrounding the organoid using bioprinting, for example [29, 74, 75]. Maturation and vascularization appear to be intrinsically linked and, thus, it is important to find a way in which these limitations can be overcome.

Another primary concern with brain organoid culture is the lack of standardisation. Due to this being such a rapidly advancing, relatively new field, there are a number of protocols used to drive organoid formation in several different ways. This lack of standardisation as well as the spontaneous nature in which brain organoids are formed, results in high levels of organoid-to-organoid variability between, and even within, experiments. This variability impacts organoid reproducibility and presents a significant challenge in comparing data and drawing conclusions based on results gained from different protocols. Overcoming the issue of variability will require fastidious quality control checks and the definition of culture conditions as much as possible. This will reduce differences between experiments and will be essential to expand brain organoids’ future applicability in epilepsy research. Additionally, standardised organoid formation protocols using a bioprinted or bioengineered standardized scaffold, to which the organoid can conform to in culture [76, 77], will also aid in reducing heterogeneity and allow the organoids to be more reproducible. Single cell sequencing with RNA tomography to assess the level of similarity between organoids, with regards to their gene expression patterns and the number of cells present, can also help in understanding the variation, or lack thereof, between the organoids used in a study [78, 79]. Spatial transcriptomics, which measures gene activity, can also be useful for researchers wanting to understand the map of gene expression within the organoid [80]. By using a combination of these methods above, the reproducibility of organoids can be improved and, thus, will help in facilitating the progression of brain organoid research.

Another limitation to the use of these organoids as a relevant model is the lack of surrounding supporting cells. Brain organoids have a limited number of cell types present within them, lacking groups of cells that are critical for brain function, with their absence reducing the physiological relevance of the samples. As noted above, a potential solution to this problem is the production of fusion organoids [67]. This allows the integration of organoids containing different cell types to produce a model which more closely mirrors the in vivo CNS microenvironment. Other solutions to this problem would be to use microspheres to create a molecular gradient or to use hydrogels to do the same. Bioprinting and bioengineering materials that mimic viscoelastic behaviours can also be used to aid this incorporation of other cell types [75, 81]. The integration of these other important cell types allows the production of a model that can better recapitulate the in vivo characteristics of the brain and allow for improved relevance when it comes to modelling neurodevelopment and neurological disorders.

Conclusions

In this review, we specifically show how brain organoids have been used in a variety of studies to understand the underlying mechanisms behind epileptogenesis and neuronal hyperexcitability, as well as revealing their success in testing the effectiveness of novel and existing anti-epileptic drugs and possible therapeutic targets. Although there are relatively few epilepsy studies that have made use of brain organoids, and they have some prominent limitations, the results generated thus far indicate that they are a model which shows great promise to be a cornerstone of epilepsy research in the future. When used in combination with other animal models and 2D culture methods, brain organoids present themselves as a powerful tool in understating epileptogenesis, with the ultimate goal of helping those suffering from epilepsy to better understand and treat their condition.

Abbreviations

CNS

Central nervous system

hiPSCs

Human induced pluripotent stem cells

hESCs

Human embryonic stem cells

EBs

Embryoid bodies

CDKL5

Cyclin-dependent-kinase-like 5

CDD

CDKL5 deficiency disorder

WWOX

WW domain–containing oxidoreductase

WOREE

WWOX-related epileptic encephalopathies

SCAR12 syndrome

Spinocerebellar ataxia type 12

DEE

Developmental and epileptic encephalopathies

GABA

Gamma-aminobutyric acid

OGD

Oxygen-glucose deprivation

MEA

Multielectrode array

AEDs

Antiepileptic drugs

Cx + GE

Cerebral cortex-ganglionic eminence

Shh

Sonic Hedgehog pathway

SOSR-Cos

Single-rosette spheroid cerebral organoids

PCE

Protocadherin-19 clustering epilepsy

Author contributions

MG and RB conceived and designed the study. RB performed the literature review and RB and AR wrote the first draft of the manuscript. All authors contributed to the reading, editing and approval of the final manuscript.

Funding

This work was supported by supported by the South African National Research Foundation (M.G., Competitive Support for Unrated Researchers - SRUG200320510242), The South African Medical Research Council (M.G., Self-Initiated Research Grant), The Council of Scientific & Industrial Research and the Department of Science and Innovation (RB and AR, CSIR-DSI Inter-bursary Support Programme (IBS)), and the University of Cape Town.

(MG, Research Development Grant, Postgraduate Research Training Grant, Enabling Grant Seeker Excellence Awards).

Data availability

All data generated or analysed during this study are included in this published article (and its supplementary information files).

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no conflict of interest.