Introduction

Epilepsy, a neurological disorder characterized by recurrent and unprovoked seizures, is a result of anomalous neural activities in the brain (Beghi et al. 2015; Beghi 2020). Globally, this disorder currently affects an estimated 65 million individuals (Akyuz et al. 2021). Understanding the microscopic mechanisms underlying seizures is crucial for effective epilepsy treatment, although the etiology of epilepsy is notably complex. It encompasses a vast array of modifications in neural circuits, which span from molecular alterations to network-level changes (Laxpati et al. 2014). Neural circuits are intricate connections formed by neurons of diverse properties and functions within the brain (Zhang et al. 2021b; Lerner et al. 2016). Neurons with distinct properties and functions establish complex connections in various forms, constituting neural circuits and neural networks at different levels (Malezieux et al. 2023). These activities take various forms, such as series, parallel, feedforward and feedback (positive or negative) (Sharpee et al. 2016). In recent years, the emergence of advanced techniques, such as viral tracing, optogenetics, and chemogenetics, has facilitated a deeper understanding of the intricate neural networks implicated in epilepsy (** neural circuits, which is due to their ability to replicate within infected neurons and to label hierarchical neural circuits through self-amplification (Kuypers and Ugolini 1990; Martin and Dolivo 1983). Subsequently, the emergence of an attenuated strain of PRV demonstrated highly specific retrograde transneuronal infections (Enquist 2002; Loewy 1998). Further refinements to the Bartha PRV have resulted in a potent tool for retrograde viral tracing, lending itself to the analysis of neural circuits in non-primate species (Jia et al. 2019). Genetic modifications to the VSV and HSV have also enhanced their utility in neural circuits studies (van den Pol et al. 2009; Beier et al. 2011, 2013). In recent years, advancements in the rabies virus modification techniques have made it an invaluable tool, capable of expressing numerous essential genes to correlate circuits to function (Sun et al. 2019c; Suzuki et al. 2019).

Optogenetics and chemogenetics are two modulatable techniques with extensive applications in manipulating neural circuits. Their widespread acceptance and application in neural circuit studies have been instrumental in our current understanding of the neural circuits involved in epilepsy. The first chemogenetic approach was demonstrated in 1911, with its great potential presented (Strader et al. 1991). Then, the second-generation tools called receptors activated exclusively by synthetic ligands, engineered from native G-protein-coupled receptors, were described (Coward et al. 1998; Chang et al. 2007). Since designer receptors exclusively activated by designer drugs (DREADDs) were first pointed out in 2007, DREADDs-based approaches have been widely adopted and improved (Armbruster et al. 2007; Vardy et al. 2015). In parallel to the development of DREADDs, ion channel-based approaches were also developed (Lerchner et al. 1998; Nietz et al. 2022). In conclusion, calcium imaging is instrumental in identifying the network nodes outside the seizure locus that are responsible for seizure initiation, propagation, and termination. It plays a significant role in the detection of calcium ions in organisms at cellular and even subcellular structure levels, providing high spatial and temporal resolution.

In Vivo Microdialysis

In vivo microdialysis serves as an essential technique for sampling the immediate vicinity of the catheter tip to identify abnormal concentrations of specific molecules in the extracellular milieu of the brain (Zestos et al. 2019). These abnormal concentrations can provide insights into the etiology of numerous brain diseases, including Huntington's disease, Parkinson's disease, Alzheimer's disease, schizophrenia, traumatic brain injury, stroke, depression, and epilepsy (Di Giovanni et al. 2009; Zestos et al. 2019; Todd and Butterworth 2001; Bjorkli et al. 2021; McAdoo and Wu 2008) (Table 4).

Table 4 In vivo microdialysis in the studies of neural circuits in epilepsy
Full size table

In vivo microdialysis has emerged as an essential tool in revealing key aspects of epilepsy, particularly in showing how treatment can influence seizure activity (Zestos et al. 2019). This technique has been instrumental in tracing changes in endogenous substances, such as excitatory and inhibitory neurotransmitters and their metabolites, at various stages: before, during, and after seizures. Luna-Munguia et al. integrated state-of-the-art microdialysis techniques with an innovative model of ictogenesis, enabling simultaneous monitoring of 24 molecules during ictogenesis to discern the chemical shifts associated with epileptogenesis and ictogenesis (Luna-Munguia et al. 2019). Similarly, Fukuyama et al. utilized multiprobe microdialysis in conjunction with ultra-high-performance liquid chromatography to examine transmission abnormalities of L-glutamate and GABA in the thalamocortical pathway of transgenic rats bearing rat S286L-mutant Chrna4 (S286L-TG) (Fukuyama et al. 2020b). In a separate study, the same team implanted a concentric direct-insertion type dialysis probe in the orbitofrontal cortex of rats after post-AMPA-evoked stimulation (Fukuyama et al. 2020a). Dhaher et al. merged brain microdialysis with mass spectrometry to investigate seizure initiation, seizure propagation, and extracellular brain levels of glutamate and GABA (Dhaher et al. 2021). Besides, Shimoda et al. employed a microdialysis probe inserted into the CA3 hippocampus via a guide cannula, preserving collected samples using liquid chromatography-tandem mass spectrometry (Shimoda et al. 2022).

The main limitations of in vivo microdialysis are the size of the probe and the damage to the tissue caused by inserting the probe. Besides, since certain neurotransmitters are low in brain dialysate, samples must be taken at regular intervals, a time scale which is far removed from the time scale of neuronal events (Holmgaard et al. 2010; Zestos et al. 2019). For in vivo neurochemical monitoring, microdialysis is often paired with chromatographic, electrophoretic, or enzymatic assays (Tan et al. 2021). Of these, liquid chromatography, mass spectrometry, and their tandem applications are most frequently employed for a detailed examination of in vivo neurochemistry during seizures (Kennedy 2013). Given the challenges related to optogenetic neurochemical monitoring, such as the requirement for high sampling rate analytical methods and in vivo neurochemical monitoring during optogenetic neuronal stimulation, the fusion of optogenetics with microdialysis sampling might offer an enlightening approach to measure neurochemical alterations before, during, and after controlled epileptic seizures (Parrot et al. 2015).

Fiber Photometry

In recent years, fiber photometry has become a popular technique for in vivo neural recording, favored for its simplicity and specificity (Legaria et al. 2022). Its primary aim is to elucidate the real-time functions of distinct neural pathways in mammalian behavior by optically recording natural neural activity in genetically specified and connectivity-defined projections (Zhang et al. 2021a). The sensitivity of this recording technique enables the in vivo direct measurement of coordinated neuronal afferent activities projecting to a specific downstream target deep within the brain of a mammal engaged in behavior. This is achieved by the application of genetically encoded Ca2+ indicators and advanced photometric devices (Legaria et al. 2022). Table 5 summarizes the applications of Fiber photometry techniques in epilepsy neural circuits in recent years.

Table 5 Fiber photometry in the studies of neural circuits in epilepsy
Full size table

Utilizing the power of fiber photometry-based calcium recording, Zhang et al. analyzed the relationship between epileptic seizures and brain calcium activities in freely behaving mice, presenting a simple procedure with minimal damage to brain tissues (Zhang et al. 2019b). Additionally, Cheng et al. integrated optogenetics and DBS with fiber photometry, using a fiber photometry system (Nan**g Thinkertech), during hippocampal kindling (Chen et al. 2020). They examined the role of dorsal raphe and its 5-HTergic neurons in epilepsy by bandpass filtering and collecting GCaMP fluorescence with a photomultiplier tube. In a separate study, researchers combined fiber photometry with c-fos map** to identify numerous hyperactivated CR neurons in the PIL. This approach helped to investigate the causal roles of these neurons in hippocampal seizures (Qi et al. 2022). To further illustrate the utility of fiber photometry, Xu et al. combined in vitro c-fos staining with in vivo fiber photometry recordings to monitor activity changes in the paraventricular thalamic nucleus (Xu et al. 2022). Besides, Murphy et al. pioneered a new tool termed the Fiber Photometry Coupled Focused Ultrasound System (PhoCUS), which allows for simultaneous monitoring of focused ultrasound effects on neural activities of subcortical genetically targeted cell types in freely behaving animals (Murphy et al. 2022). This advancement provides a new way to investigate focused ultrasound effects on specific cell types in deep brain regions.

Evidently, fiber photometry offers a promising method for advancing studies into the pathogenesis of epilepsy and epileptiform signal transmission. This is attributable to its exceptional sensitivity and high-resolution detection capabilities, especially regarding calcium signal detection in multiple brain regions. However, similar to electrophysiology, fiber photometry faces the problems of low resolution and limited range of brain area monitoring, and its difficulty in controlling the depth of captured signals, which are worthy of subsequent research and improvement (Mineur and Picciotto 2023).

Virus Tracing Technology

Due to the collaborative efforts across various disciplines, particularly the combined progress in molecular biology, virology, microscopy, computer science, and genetics, effective advancements have been achieved in the methodological aspects of neuron tracing. Viral tracing techniques are a hallmark product of this progress. In the past twenty years, viral tracing techniques have garnered a surge of interest in neural circuitry studies, especially in epilepsy studies (Table 6). They exhibit a unique ability to deliver tracers in a cell-type-specific and circuit-selective fashion, employing cis-, retrograde, and trans-synaptic methods (Liu et al. 2022).

Table 6 Virus tracing technology in the studies of neural circuits in epilepsy
Full size table

Innovative studies have applied rabies tracing techniques to explore neural circuits relevant to epilepsy with some success. Studer et al. utilized rabies virus-mediated retrograde synaptic tracing to investigate a mouse model of atonic epilepsy (Studer et al. 2022). Their results showed that cortical neurons, located in both the superficial and deep layers of the whisker primary somatosensory cortex, demonstrated significantly increased and abnormal connectivity in a genetic model of anhedonic epilepsy. This suggests that the enhanced structural connectivity patterns of these neurons might be a crucial pathological basis for increased neuronal synchronization and the generation of anhedonic seizures. In sum, this research offers new insights into the relationship between neural hypersynchronization and abnormal structural connectivity in neuronal networks during epileptic seizures.

Rabies virus-mediated presumptive retrograde trans-synaptic tracing is frequently combined with other tracing techniques for improved performance, such as alternative viral tracing methods, optogenetics, and chemogenetics (Islam et al. 2022; Kato and Kobayashi 2020; Wang et al. 2020). In an experimental rat model of medial temporal lobe epilepsy (mTLE) induced by trichothecene, Du et al. implemented a dual viral tracing strategy, which combined retroviral birth dating with rabies virus-mediated presumptive retrograde trans-synaptic tracing (Du et al. 2017). They observed substantial remodeling of the hippocampal circuit following epileptogenic injury, causing DGCs to generate significant excitatory monosynaptic inputs, such as local recurrence and widespread feedback loops. A similar outcome was achieved by Zhou et al., who integrated rabies virus tracing techniques with chemogenetic approaches (Zhou et al. 2019). Future research could consider utilizing other tracing techniques or examining unexplored subtypes of interneurons.

Recently, Wang et al. integrated optogenetics and chemogenetics with a combination of in vivo and in vitro electrophysiology, as well as retrograde rabies virus tracing (Wang et al. 2020). This innovative combination helped to elucidate a direct septo-hippocampal cholinergic circuit, which efficiently mitigates seizures by instigating growth inhibitor inhibition in an animal model of TLE. This study improves our understandings of alterations in seizure circuits and their precise spatiotemporal controls. Despite these advancements, few studies have documented brain circuits activities during seizures through recombinant adenoviral methodologies. Currently, lentiviruses and rabies viruses are widely used in the tracing of neural circuits (Kato and Kobayashi 2020; Deng et al. 2022). It is crucial to choose different viral vectors based on specific research objectives.

There are also several limitations in virus tracing technologies. Due to the specificity of the vector, it often faces biosafety issues such as neurotoxicity (Sun et al. 2021; Christenson Wick et al. 2017). Chen et al. investigated the complex circuitry of TLE, suggesting that the long-range SNr-PF inhibitory circuit is involved in regulating seizures in TLE and the deactivation of this circuit could lessen seizure severity (Chen et al. 2020). In a comparative study of pathological changes in the medial septum's (MS) hippocampal circuits between TLE patients and healthy individuals, Wang et al. discovered that direct MS-hippocampal cholinergic projections mediated antiepileptic effects by preferentially targeting hippocampal GABAergic neurons, underscoring the specific role of the MS-hippocampal cholinergic circuit in TLE (Wang et al. 2020). Additionally, aberrant sparse activity of hippocampal dentate gyrus (DG) granule cells (GCs) has been found to contribute to TLE, with recent optogenetic evidence indicating that the axons of glutamatergic mossy cells play a pivotal role in this process (Botterill et al. 2019). These studies hold significant potential for precisely targeting epilepsy.

Furthering the integration of optogenetics with viral tracing techniques, Sere et al. investigated the role of lateral hypothalamic neurons (LH) in dyscalculia epilepsy (Sere et al. 2021). The study illustrated that LH neurons extend to the dorsal raphe nucleus (DRN) in the brainstem, with a subset being GABAergic. It confirmed the presence of long-distance subcortical modulation of serotonergic neuromodulation during anhedonic seizures, enhancing our understanding of seizure complexities. It is known that pulmonary portal mossy cells can influence hippocampal circuits function, and substantial loss of these cells is associated with alterations in the same circuits during epileptic states. Optogenetics has been instrumental in delineating this involved microcircuitry (Butler et al. 2022). Seeking to clarify how abnormal neural circuits contribute to epilepsy, Yang et al. examined circuit mechanisms in a mouse model of focal cortical malformations through optogenetic stimulation (Yang et al. 2021). Their findings revealed site-specific and cell-type-specific synaptic reorganization within epileptic cortical circuits, highlighting the need for more extensive research to understand the prevalence and overall landscape of these abnormal circuits. Besides, recent work has demonstrated that calretinin (CR) neurons in the posterior intralaminar thalamic nucleus (PIL) exhibit varying influences on hippocampal seizures via distinct downstream circuits, adding to our understanding of the PIL-lateral amygdala CR circuit (Qi et al. 2022).

Optogenetics offers the ability to accurately activate or inhibit-specific neurons, serving as a research tool and potential therapeutic strategy for managing aberrant neuronal activities. GABAergic neurons have been proposed as novel targets for optogenetic modulation of temporal lobe seizures. As the scope of animal studies and clinical treatment broadens, optogenetics holds considerable promise in elucidating epileptogenesis, seizure onset, and pathological circuits, which could potentially open avenues for alternative therapeutic strategies such as gene therapy. The study of neuronal interactions during seizures with the aid of optogenetics has garnered increasing interest and is expected to become a central area of investigation in epilepsy research.

Optogenetics also faces some dilemmas. The expression of photosensitive proteins is not uniform across neuronal populations, which can lead to heterogeneity in the magnitude and scope of optogenetic manipulation (Melonakos et al. 2020). In addition, the simultaneous action of a large range of light stimuli on a population of neurons may cause circuits to show non-physiological patterns of activity (Peralvárez-Marín and Garriga 2016). And because of the complexity of neural circuits, optogenetics is more difficult to apply to the primate brain, which limits its possible wide application (Merlin and Vidyasagar 2023).

Chemogenetics

Similar to optogenetics, chemogenetics is widely applied in neuroscience for the modulation of specific neurons or neural circuits. Chemogenetics is a method which utilizes gene technology and artificially designed drugs to alter the function of neurons (Poth et al. 2021; Magnus et al. 2019). It is commonly employed to explore the function and activity dependence of neurons. This technique involves the introduction of engineered receptors or channels, which selectively respond to synthetic ligands, thereby modulating cellular signaling pathways. Table 8 summarizes the recent applications of chemogenetics in the neural circuits of epilepsy. Chemogenetics can either be direct, through the utilization of synthetic ligand-gated ion channels, or indirect, by leveraging DREADDs, tools derived from G-protein-coupled receptors (Oyrer et al. 2018). DREADDs have gained popularity as tools of chemogenetics in studying the neural circuits of epilepsy due to their capability to either inhibit or excite neurons. This is achieved through the expression of designer receptors which respond to exogenous compounds such as clozapine N-oxide or olanzapine.

Table 8 Chemogenetics in the studies of neural circuits in epilepsy
Full size table

Chemogenetic tools have proven instrumental in the selective activation or silencing of pathways in a specific population of neurons. It has greatly facilitated the investigation of anatomically constrained neural circuits in epilepsy and the development of high-resolution circuit maps associated with seizures. Chemogenetics has illuminated the circuit-based mechanisms underlying seizures. For instance, Kahn et al. used DREADDs to manipulate the activity of dentate gyrus cells (DGCs) to target the aggregate output of the dentate gyrus, thereby demonstrating the importance of fine-tuning activity within this region for cognitive performance (Kahn et al. 2019). In addition, Zhou et al. have underscored the crucial role of newborn DGCs in the hippocampus in the formation of epileptic neural circuits and the initiation of spontaneous recurrent seizures (Zhou et al. 2019). Through the application of chemogenetic methodologies, circuits such as the nigra-parafascicular disinhibitory circuit, which regulates seizures in TLE, and the dHPC dorsal lateral septum circuit that impairs risk assessment behavior, have been identified and evaluated (Chen et al. 2020; Cao et al. 2022). Projections from the mediodorsal nucleus to the prelimbic cortex have also been implicated in the propagation of amygdala-kindled seizures (Wicker et al. 2022). An increasing number of studies are leveraging chemogenetics for the specific activation and inhibition of neurons associated with particular circuits, thus assessing the impact of chemogenetic interventions on epileptic activity (Berglind et al. 2018). Furthermore, develo** novel DREADDs ligands and investigating new mouse models are subjects of current research, and improvements in these areas are anticipated (Desloovere et al. 2022).

Interneurons have garnered attention as pivotal components for understanding the neural circuits involved in epilepsy, with particular focus on the three main interneuron populations in the rodent hippocampus: parvalbumin (PV), somatostatin (SST), and vasoactive intestinal peptide expressing interneurons (Miles et al. 1996; Paz and Huguenard 2015; Lovett-Barron et al. 2012). Cǎlin et al. utilized excitatory DREADDs to activate different interneuron subtypes, finding that their ability to increase postsynaptic inhibition in principal neurons and reduce epileptiform synchronization in neuronal networks varies. However, all these subtypes may serve as effective anticonvulsant strategies (Călin et al. 2018). Besides, Panthi and Leitch focused on the activation of PV + interneurons using DREADDs, and they administered clozapine N-oxide to activate these interneurons (Panthi and Leitch 2021). Their study suggested that FFI of PV + interneurons within cortico-thalamocortical microcircuits could inspire new advancements in anti-absence seizure treatment. A separate study emphasized the role of DG PV-INs, proposing that the inhibitory function of PV-INs could be a promising therapeutic target for seizure modulation (Mattis et al. 2022). Notably, GqDREADD-mediated chemogenetic activation of wild-type SST interneurons could lead to an increase in spontaneous excitability and susceptibility to depolarization block, thereby contributing directly to seizures in vivo and highlighting the substantial role of SST interneurons in seizure events (Wengert et al. 2021).

Moreover, the field of chemogenetics is proving to be groundbreaking in making strides toward clinical translation in the treatment of epilepsy. Gene therapy, which enables the manipulation of neuronal activities in the epileptic focus while simultaneously preserving function, appears as a promising alternative to traditional pharmacological methods and surgical interventions for epilepsy. By expressing proteins that can be regulated by drugs, chemogenetics can serve as a titratable therapy that could potentially overcome challenges associated with traditional therapies that are limited by fixed gene dose, expression of that gene, and the distribution of transfected cells (Kullmann et al. 2014; Walker and Kullmann 2020). Zhang and Wang have summarized and elucidated the benefits of chemogenetics for gene therapy and several studies have demonstrated the potential effectiveness of chemogenetics in epilepsy treatment, thereby supporting its prospective clinical translation in the future (Zhang and Wang 2021).

Chemogenetics could result in minimal damages to animals. After viral expression in the brain, each experiment only requires the injection of drugs into muscles or veins to inhibit or excite a specific brain region, making it highly suitable for translational medical research (Song et al. 2022). However, compared to optogenetics, chemogenetics also has its drawbacks, particularly in terms of lower temporal precision. The effects of drugs persist for several hours, during which neurons remain in an abnormal state. Cognitive processes in the brain often change rapidly, with multiple cognitive events occurring within seconds or even one second. Chemogenetics may struggle to separate these transient and sequentially occurring cognitive processes (Raper and Galvan 2022).

In Vitro Models

In response to the growing sophistication of tools for the exploration of neural circuits in epilepsy, there is an emergent need for more advanced in vitro models. These models could serve to address current limitations and bolster biomedical research, with a particular focus on the discovery of antiepileptic drugs (Quadrato et al. 2017; Elder et al. 2022). Traditional murine models have demonstrated efficacy in elucidating the impact of epilepsy-associated genes on neural cells and circuits. However, they are unsuitable for high-throughput screening procedures and exhibit dissimilarities in brain development compared to their human counterparts (Silbereis et al. 2016; Jiang et al. 2018b). In efforts to overcome these challenges, human stem cell-derived models have been introduced. These models include neurons derived from iPSCs and organoids cultured from patient cells, both of which offer a patient-specific genetic background and human-specific cell types.

iPSC-Derived Neurons

Over recent years, iPSCs have taken a prominent role in epilepsy research, given their inherent pluripotency which lends itself well to modeling neurological disorders such as epilepsy (Shi et al. 2012). It has been found that iPSCs could reproduce the genetic mutations and background of the affected brain cells, thus making them suitable for modeling these disorders (Bhargava et al. 2022). Following the initial success in generating iPSCs from mouse fibroblasts, significant advancements have facilitated their application in epilepsy pathogenesis studies (Karagiannis et al. 2019).

A variety of genetic mutations, especially in different ion channels, play a role in the development of epilepsy. These mutations can affect the excitatory/inhibitory (E/I) balance and the stability of neural circuits. Key channels involved include voltage-gated sodium, potassium, and calcium channels, along with ligand-gated glutamatergic and GABAergic receptors (Oyrer et al. 2018). However, few of these mutations have been examined in humanized in vitro models. With the application of epilepsy patient-derived iPSCs to generate both excitatory and inhibitory neurons, voltage-gated sodium channels (Nav), including Nav1.1, Nav1.2, and Nav1.6 (encoded by SCN1A, SCN2A, and SCN8A, respectively), which have been associated with epilepsy, have been examined for their essential role in the depolarizing phase of the action potential (Child and Benarroch 2014). While most studies are focusing primarily on SCN1A and only investigating Nav1.1 in vitro, Tidball et al. have developed iPSC lines from epilepsy patients with SCN2A and SCN8A LOF mutations for prospective studies (Kim et al. 2018; Tidball et al. 2017). Mutations in voltage-gated potassium channels, such as Kv7.2 and KNa1.1, have also been modeled in vitro (Simkin and Kiskinis 2018).

It is understood that the hippocampus' homeostatic set point is regulated by dihydroorotate dehydrogenase. However, maladaptive changes can readily disrupt this homeostatic plasticity. Mutations in ion channels and components of the mTOR pathway have surfaced as a primary area of focus for neural circuits research in epilepsy (LaSarge et al. 2021; Weng et al. 2022). The dynamic interactions between these components warrant further investigation through humanized in vitro systems. These systems offer a promising platform to identify homeostatic alterations within neural circuits, which could contribute to epileptogenesis.

There are also several limitations of iPSCs technology. For complex, multigenic epilepsy types, iPSCs technology is difficult to model effectively. In addition, the overall low reprogramming efficiency and the high cost used to produce and characterize each cell line discourage their application in experiments (Huangfu et al. 2008; Sahlgren Bendtsen and Hall 2023). More basic research is still needed to develop this technology.

Organoids

2D cultures of iPSC-derived neurons can offer a valuable understanding of molecular mechanisms at the individual neuron level (Heider et al. 2021). However, they fall short in enabling investigation into neurogenesis and neuronal circuits. Moreover, they lack the 3D architecture characteristic of the brain and exhibit immature electrophysiological properties (Malik and Rao 2013). Organoid models effectively address these constraints, particularly in diseases associated with dysfunctional neural circuits, such as epilepsy (Andrews and Kriegstein 2022). Advances in human iPSC techniques have facilitated the development of 3D brain models, including brain organoids and cerebral organoids (Swingler et al. 2023). These models have often been employed for in vitro neuro-developmental research, but their potential in the context of epilepsy, especially for studying neural circuits, has been relatively underexplored (Pranty et al. 2022).

Birey et al. employed brain cortical spheroids (a specific type of cerebral organoid) to examine the developmental features of neurodevelopmental disorders, including epilepsy (Birey et al. 2017). Another study conducted by Blair et al. created cortical organoids from genetically engineered stem cell lines to investigate tuberous sclerosis (TS) complex (Blair et al. 2018). Additionally, Litwa further discussed the ability of brain organoids to reveal TS pathology, unravel disease mechanisms, and illuminate potential contributions of neurodevelopmental alterations to subsequent age-related neurodegeneration (Litwa 2022). Besides, Sun et al. developed 3D cortical organoids, which followed a normal developmental trajectory, to investigate whether functional deficits are relevant to the progression of Angelman syndrome and model epilepsy susceptibility in this syndrome (Sun et al. 2019a). Variations in neuronal network dynamics in numerous cells were monitored in both wild-type and knockout cortical organoids. In a study by Samarasinghe et al., calcium sensor imaging and extracellular recording techniques were used to analyze the functions of brain organoids at the network-level (Samarasinghe et al. 2021). Their research indicated that the fusion organoid system is amenable to detailed cellular and circuit analyses, providing insights into modeling human neurological disorders. Given the absence of in vivo connectivity, which hinders the integration with other circuits controlling behaviors, Revah et al. transplanted human stem cell-derived cortical organoids into the somatosensory cortex of newborn athymic rats (Revah et al. 2022). Their results demonstrated that these organoids were able to develop mature cell types and integrate into sensory and motivation-related circuits.

Although organoids have developed rapidly in recent years, for brain organoids, their development has not been smooth due to the lack of an in vivo brain tissue microenvironment, neuronal circuits, blood vessels, and immune system (Cakir et al. 2020). This collection of techniques facilitated a thorough understanding of the MS-hippocampus cholinergic circuit's role in TLE. The incorporation of imaging, neuronal manipulation, and circuit tracing allows researchers to examine pathological alterations, manipulate-specific neurons, and trace circuit connections, ultimately resulting in solid conclusions regarding the circuit's functionality and its prospective therapeutic significance in epilepsy. In a similar vein, Sere et al. investigated the role of the LH in absence epilepsy through the utilization of a blend of viral tract tracing, optogenetics, and both in vitro and in vivo electrophysiology (Sere et al. 2021). These techniques facilitated the characterization of neuronal activity within the LH and its relationship with the brainstem dorsal raphe nucleus, a region associated with the serotonergic system. The researchers discovered that LH neurons project to the DRN and their activity correlates with spike-wave discharges during absence seizures. This finding suggests that the LH, an area of the brain involved in autonomic regulation and heavily innervating the DRN, has a role in absence seizures and related comorbidities. It supports the hypothesis of a long-range subcortical regulation of serotonergic neuromodulation during such seizures.

However, the more combined techniques there are, the inevitably higher costs and more complex integration schemes will be faced. In summary, as the neural circuits studied become increasingly complex, the problems inherent in single techniques cannot be ignored, and the combination of multiple techniques is inevitably becoming a mainstream trend.

Drug Therapy Targeting Neural Circuits in Epilepsy

GABA deficiency, excessive serotonin activity, hyperactive dopamine, and glutamate excitotoxicity could enhance the occurrence of epilepsy. Neurons containing these neurotransmitters constitute the main neural circuits of epilepsy (Werner and Coveñas 2019). In recent years, more and more drugs targeting neural circuits have been developed for the treatment of epilepsy patients (Fig. 3). These drugs usually work by acting on specific neuronal cells, either exciting or inhibiting the release of synaptic vesicles, preventing excessive excitation or inhibition of nerve cells, and ultimately regulating the relevant axis to suppress epileptic seizures (Huang et al. 2023; Gunn and Baram 2017; Mao et al. 2022). Additionally, the latest research, such as investigating neuronal connection patterns through AI algorithms and using them as the basis for functional operation, was not discussed separately due to limited literature (Gao et al. 2023; Cao et al. 2023; Zhang et al. 2021b). Undoubtedly, these technologies are also very significant for the study of epilepsy. In conclusion, the array of methods utilized to explore neural circuitry in epilepsy has resulted in significant insights into the disorder's pathophysiology. As we continue to refine and expand these techniques, the prospect of identifying new therapeutic targets and devising more effective, individualized treatment for epilepsy patients becomes ever more attainable. Through continued interdisciplinary collaboration and technological innovation, we anticipate substantial progress in the treatment and clinical management of epilepsy.