Among the primary genetic tools used in Drosophila are binary expression systems, e.g., the GAL4/UAS, LexA/LexAop or Q-System (Brand and Perrimon, 1993; Lai and Lee, 2006; Riabinina and Potter, 2016). As the name implies, two components are needed to induce gene expression: a driver of gene expression and a sequence whose expression is controlled by the driver. The driver usually consists of a transcription activator, which is expressed under a tissue-specific promoter. This activator then binds a specific genetic sequence and drives expression of a gene linked to this sequence. In the most frequently used binary expression system, the GAL4/UAS system, expression of a UAS-transgene is achieved through the exogenous (yeast) transcription activator protein GAL4 (Figure 3). The UAS-line harbors a gene/sequence of choice downstream of GAL4 binding sites, called upstream activating sequence (UAS). The UAS-controlled sequence is not expressed in the absence of GAL4. Expression of the UAS-controlled sequence is achieved by crossing a GAL4-driver line with a UAS-line. In the progeny, GAL4 will bind to UAS and facilitate expression of the target sequence downstream of UAS. The GAL4-driver lines were generated as random integration enhancer traps. This way, expression of GAL4 is controlled by an endogenous promoter, which defines the expression domain and allows restricting expression to a region of interest. To date, a plethora of characterized driver lines (> 10,000) are available at aforementioned stock centers and many transgene carrying UAS-lines are available there as well. Some of the most frequently used GAL4-driver lines can be found on FlyBase³. Drivers commonly used in the study of the nervous system are, for example, elav-GAL4 or nsyb-GAL4, which allow for pan-neuronal expression of GAL4. Further examples include VGlut-GAL4 and ChAT-GAL4 for confining GAL4 expression to glutamatergic and cholinergic neurons, respectively. The expression of GAL4 follows the expression pattern of the associated gene, which might vary during development of the fly. This should be considered when designing experiments.

Binary expression systems can be used, for example, to overexpress cDNA from a human disease-related gene or the corresponding Drosophila orthologue (Jenett et al., 2012; Perkins et al., 2015). This cDNA can either be wild-type or a disease-associated variant. If a disease phenotype is observed after expression of the disease-associated variant, but not after expression of the wild-type, the variant is likely a toxic GOF (Prüssing et al., 2013). This approach has been used to classify para^bss1 as a GOF variant (Parker et al., 2011b) or to replicate a severe DEE syndrome by overexpression of RhoBTB (Straub et al., 2018). Another possible application of binary expression systems is the induction of RNAi. If a LOF variant is the hypothesis for the disease, RNAi can be used to model the loss of protein function. Expression of inverted repeats (ir) or short hairpins (sh) causes formation of double stranded RNA (dsRNA). If the dsRNA is complementary to endogenous mRNA, this leads to the degradation of target mRNA and a decrease in protein levels (Kaya-Copur and Schnorrer, 2016; Heigwer et al., 2018). The strength of RNAi is influenced by several factors including the levels of interfering RNA, which in turn depends on the activity of the GAL4/UAS system.

To understand the effect of a variant on the phenotype, a rescue experiment can be performed (Ecovoiu et al., 2022). Here, a disease-related phenotype is alleviated by reintroducing a functional allele. For instance, in the case of an epilepsy-related variant in the gene eas, expressing wild-type eas alleviated the seizure phenotype (Kroll and Tanouye, 2013). Besides already established or newly found epilepsy-related variants in genes, RNAi or deficiency lines can be used as a background for rescue experiments. If a wild-type allele rescues the disease phenotype and a variant does not, this provides compelling evidence for a LOF variant.

Though the GAL4/UAS system is a powerful tool, it requires several considerations. The timing of gene expression is not variable and expression strength is generally elevated at higher temperatures, as GAL4 activity increases. To allow for more precise temporal control over gene expression, inducible GAL4 systems can be employed (Supplementary Figure 1). Additional expression of GAL80 can prevent target gene transcription by inhibition of GAL4 (Suster et al., 2004). Moreover, the temperature-sensitive variant GAL80^ts is only active at lower temperatures (Ma and Ptashne, 1987; McGuire et al., 2004; del Valle Rodriguez et al., 2011). This way, temporal control of GAL4 activity is achieved by shifting flies to higher temperatures to inactivate GAL80^ts. Instead of depending on temperature shifts, the GeneSwitch-GAL4 system depends on the presence of the steroid mifepristone to activate GAL4, which can be added to the standard feeding medium of flies (Osterwalder et al., 2001; del Valle Rodriguez et al., 2011). To allow for better spatial control of GAL4 expression, the split-GAL4 system can be utilized. Here, two different promoters drive expression of a truncated GAL4. Only in cells where both promoters are active, the complementary GAL4 peptides assemble a functional GAL4 (Luan et al., 2020). Even with more specific tools available, these versions of binary expression systems do not fully capture expression profiles of genes of interest, though their easy accessibility makes them useful in initial investigations and for large screens.

The phenotypic outcome of a given epilepsy-linked variant might be dependent on gene dosage or timing. Here, the mentioned binary expression systems might fall short. To allow better control over expression levels, an endogenous promoter can be hijacked. The transgene can then be expressed in the same pattern of a gene of interest. This can be achieved by utilizing a Minos-mediated integration cassette (MiMIC; Venken et al., 2011). The transposable element Minos randomly integrates in the genome and harbors a pair of attP landing sites. Via the bacteriophage-derived ΦC31 integrase, recombination mediated cassette exchange (RMCE) can be performed. Ultimately, this creates a system with defined integration sites in the Drosophila genome, in which a desired genetic sequence can be integrated. Different alterations can be achieved based on the position of the insertion. To drive gene expression, the RMCE site can be placed under an endogenous promoter. By targeting an intron, it can be used to tag a protein with a fluorescent protein. As this system is based on random integrations, the efficiency in creating new lines diminishes with each new line generated.

Besides Minos, CRISPR-based strategies (CRIMIC) can also be used to introduce RMCE sites into the genome. Here, integration can be targeted to any position in the genetic sequence accessible to CRISPR/Cas9 modifications. For introns accessible with either method, T2A-GAL4 (“Trojan exon”) can be used to express GAL4 under a promoter of the gene, whose own expression is simultaneously suspended (Diao et al., 2015; Lee et al., 2018). GAL4 can then drive the expression of any UAS construct, while its own expression is controlled by the promoter of the disabled gene. Using this system, different variants can be rapidly screened for their ability to restore protein function, while following endogenous expression patterns. Mutations in the human gene OGDHL cause a neurodevelopmental spectrum disease featuring epilepsy. Using T2A-GAL4, gene function could be restored by wild-type expression, but only partially by patient variants introduced into the Drosophila orthologue dOgdh (Yap et al., 2021). In a similar approach, LOF variants in TIAM1, which are associated with seizures, were introduced into Drosophila and screened for their ability to rescue a loss of the Drosophila orthologue sif (Lu et al., 2022). The lower rescuing capacity of patient variants corroborated a LOF effect. As not all genes are suitable targets for CRMIC-based integration of a trojan exon, the gene can also be replaced completely by a GAL4-expressing sequence (Kanca et al., 2022). Large libraries of flies with integration sites in different genes provide broad access to the Drosophila genome (Venken et al., 2011; Nagarkar-Jaiswal et al., 2015; Lee et al., 2018).

The most detailed analysis of protein function can be achieved by directly altering the gene itself in the fly genome. In case a human epilepsy-linked mutation is located in a conserved region, the fly gene can be mutated at the respective position to create a humanized variant. Ends-out homologous recombination allows the replacement of the wild-type gene by a mutated variant thereof (Rong and Golic, 2000). Using this technique, patient variants associated with GEFS+ (Sun et al., 2012) or DS (Schutte et al., 2014) have been introduced into the fly genome. To achieve precise genome modifications more efficiently, the CRISPR/Cas9 system has become a widely used tool for genome editing (Jinek et al., 2012; Gratz et al., 2015). In flies, it has been used to introduce patient-specific variants into the gene para, which result in either GEFS+ or DS (Roemmich et al., 2021). This allows to study patient-derived variants in a comparable genetic background and to assess, which physiological alterations might be caused by the variants.

Through these methods, genomic modification on different scales becomes feasible for many conserved genes in Drosophila. Libraries of fly strains generated by these methods provide researchers access to nearly every gene in Drosophila, while direct gene editing remains a possibility to investigate single genes in greater detail.

To induce a bang-sensitive phenotype, it has become the standard to use a laboratory “vortex mixer” to shake flies in an empty vial for ~10 s (Kuebler and Tanouye, 2000). If flies have the bang-sensitive phenotype, they display seizure-like behavior and become paralyzed, while wild-type flies remain unaffected. The time until affected flies regain a standing position or mobility correlates with the severity of the phenotype. Administration of ASMs has been shown to reduce the severity of the phenotype in some bang-sensitive mutants (Kuebler and Tanouye, 2002). This reduction could also be achieved by different mutations, which were hence termed seizure suppressors (Saras et al., 2016).

To screen for temperature-related neuronal defects, a reliable way is through a heated water bath, in which flies collected into empty vials can be placed (Burg and Wu, 2012). At a water temperature of ~40°C, seizure-like behavior and paralysis can be observed in susceptible flies as the temperature in the vial increases. Time until seizure or paralysis occurrence, as well as recovery time can be compared. Alterations to seizure susceptibility via genetic or pharmacological manipulation can lead to a more heat-resistant phenotype or a faster recovery. In Drosophila models of GEFS+ and DS, seizure-like behavior and paralysis occurred at elevated temperatures. A faster onset could be observed in the DS model, corresponding to the increased severity of the variant in patients (Sun et al., 2012; Roemmich et al., 2021). Mutations in other genes can induce non-seizure-related paralysis at elevated temperatures. A notable example is the shibire mutant shi^ts, which encodes the Drosophila Dynamin orthologue Shibire. At high temperature, the mutant Shibire protein adopts an inactive conformation. As a consequence, there is a block of synaptic vesicle scission, resulting in paralysis (Kosaka and Ikeda, 1983; Hill et al., 2001). A return to lower temperatures allows the mutant Shibire protein to adopt its normal conformation and restores function. Interestingly, overexpression of shi^ts in neurons of bang-sensitive flies leads to a suppression of mechanically induced seizures at an elevated temperature, as endocytosis becomes impaired (Kroll et al., 2015b).

These assays cover only a portion of behavioral assays in Drosophila. The behavior of flies can also be assessed in terms of activity, locomotion, memory, or social behaviors (Nichols et al., 2012). For instance, sleep disturbances and changes in locomotor activity were observed in a knock-in model of human GEFS+ (Petruccelli et al., 2015; Mituzaite et al., 2021). Also, flies deficient for sif exhibited severe climbing defects besides seizure-susceptibility (Lu et al., 2022). Follow-up investigations of a newfound phenotype with different means are therefore necessary for a comprehensive description.

Electrophysiological recordings allow a more detailed investigation of disturbed neuronal function in Drosophila. Seizure-like activity can be induced in all flies, regardless of genetic background, by direct electrostimulation of the brain (Howlett and Tanouye, 2009). To achieve this, electrodes are inserted in both hemispheres of the brain and a short, high-frequent electric stimulus is applied. Depending on the fly’s susceptibility to seizures, different voltages are required to induce a seizure. In bang-sensitive mutants, low voltages are sufficient to induce seizure-like activity, while wild-type or seizure-resistant mutants require higher voltages for seizure induction (Lee and Wu, 2002). Electrophysiological recordings can be performed at the flight or leg muscle. These muscles are innervated by the giant fiber (GF) pathway, an electrophysiologically well-characterized system (Allen et al., 2006). A seizure disturbs GF function and leads to a characteristic display of muscle potentials. While low-intensity single pulse stimulation at the brain is sufficient to activate the GF pathway and create a signal at the innervated muscle, this transmission is abolished during and shortly after a seizure. With this approach, the seizure phenotype of different genotypes can be quantitatively investigated. This includes the voltage needed to induce seizure-like activity, as well as the recovery time, until GF function is restored. Compared to behavioral assays, this method requires a more complex experimental setup and is less suitable for high-throughput applications. To investigate changes in neuronal function in a broader matter, electroretinograms (ERG) can be recorded from the compound eye of the fly (Wu et al., 2022). Here, light-invoked potentials from photoreceptors and neurons are measured at the surface of the compound eye. Changes between measured responses between different genotypes can reflect defects caused by mutations. Besides these methods, patch-clamp experiments at neurons of the adult fly or larvae are possible as a readout for the activity in individual neurons (Marley and Baines, 2011; Sun et al., 2012).

A non-invasive approach to induce seizures directly at the brain can be taken via optogenetic tools. In optogenetics, neuronal function is altered by introducing channelrhodopsins (ChR), which are light-sensitive ion channels (Zhang et al., 2007). Upon stimulation with light of a specific wavelength, the channels open and allow for the flow of anions or cations. This way, neuronal function can be directly controlled. Using the expression systems described above, ChR expression can be targeted to specific areas of the nervous system. Expressing a red-shifted ChR carrying an excitatory current in the brain of several bang-sensitive mutants allows light-activated seizure induction in these flies (Saras et al., 2017). On the other hand, activation of ChR limited to mushroom body neurons was sufficient to induce seizure-like activity in para^bss1 flies. This technique offers a way to pinpoint centers of seizure generation in the fly brain. Optogenetic manipulations can not only be utilized for direct seizure induction in adults but also to alter neuronal function of flies during development. DEEs carry a developmental component, which is closely linked to their etiology (Specchio and Curatolo, 2021). A better understanding of variants in disease-causing genes during development will provide a better understanding of disease etiology and allow for a rational design of therapeutic strategies to treat these syndromes. While embryogenesis is challenging to study in mammals, the development of Drosophila can be easily observed and manipulated. Expression and activation of ChR pan-neuronally during the development of wild-type flies has been shown to permanently alter network activity and to induce persistent seizure-like activity even in adult flies (Giachello and Baines, 2015). Further limiting these network alterations to specific, critical periods in development has also been sufficient to create this phenotype. Longitudinal administration of ASMs alleviated the seizure phenotype. Besides optogenetic approaches, limiting genetic alterations to these critical periods can lead to a seizure-phenotype as well, highlighting the importance of development-specific effects (Horne et al., 2017).

Besides ChR activation by light, gene expression can also be induced by light (Chan et al., 2015; Lee et al., 2017). Here, photosensitive proteins like cryptochrome are linked to effectors, which drive gene expression following a cascade of interactions after light exposure. As a tool combining light-sensitivity with the GAL4/UAS system, PhotoGal4 offers a light-dependent gene expression system which can be used in conjunction with established UAS-lines (de Mena and Rincon-Limas, 2020). The plant-derived phytochrome B is fused to the GAL4-DNA binding domain (GAL4-DBD), while the phytochrome-interacting factor 6 is fused to a VP16 activator domain (VP16-AD). Exposure to red-light allows for a conformational change of phytochrome B and subsequent binding to the phytochrome-interacting factor 6, which brings together the GAL4-DBD and VP16-AD and induces gene expression. Illumination with light in the far-red spectrum reverts phytochrome B to an inactive form, thereby stopping the interaction and gene expression rapidly. A drawback of this method is the requirement of an exogenous chromophore. Another tool that utilizes light-sensitive proteins fused to a GAL4-DBD or AD is ShineGal4 (di Pietro et al., 2021). Here, DBD and AD are brought together by the interaction of “magnet photoswitches”. These proteins act like magnets and heterodimerize upon exposure to blue light (Supplementary Figure 1). By utilizing the GAL4/UAS system, these systems are “backwards compatible”. This expands the possibilities of investigating gene function at precise points in time during development or in the adult fly.

To further understand network functions and alterations, neuronal activity in flies can be visualized using several strategies (DeNardo and Luo, 2017). Calcium imaging has been extensively used as a readout of neuronal activity (Grienberger and Konnerth, 2012). As calcium is an essential cellular messenger, fluctuations can be directly correlated to neuronal activity and visualized through chemical or genetically encoded calcium indicators (GECIs). Since conventional GECIs require monitoring during behavioral tasks, their use in behavioral applications becomes challenging. For example, loss of the gene cpes leads to light-induced seizure behavior, which was also visible as an increase in neuronal activity in a calcium imaging approach in the fly brain (Kunduri et al., 2018). In a simpler setup suitable for larger screens, the calcium-modulated photoactivatable ratiometric integrator (CaMPARI) has been used to label active neurons in freely moving flies (Fosque et al., 2015). CaMPARI is a calcium indicator that uses two components to display neuronal activity. Under conditions where calcium and UV light are available, CaMPARI undergoes a molecular switch and permanently changes from green to red fluorescence. Using UV light only during a defined experimental stimulus, neurons activated in this temporal window can be labeled for functional mapping of neural networks (Zolnik et al., 2017). The technique has been used in Drosophila to mark acid-sensing neurons in flies, while they were allowed to move freely in an environment containing either acid or neutralized acid (Edwards et al., 2020). In the context of seizure-like activity, this method might help to further investigate which neuronal networks are crucial during seizure generation. The role of ASMs and other factors modulating seizure-like activity can also be assessed in a visual manner via calcium imaging. Preparations of the larval nervous system provided another possibility for large-scale screening of ASMs, as they influenced the spatial–temporal patterns of waves of calcium activity between segments of the ventral nerve chord (Streit et al., 2016).

Other approaches to trace calcium in neurons are based on transcriptional activators. In the presence of calcium, they drive expression of downstream targets like GFP, which then accumulates in recently active neurons. In the CaLexA system (calcium-dependent nuclear import of Lex A), the calcium-responsive transcription factor NFAT (nuclear factor of activated T cells) is fused to LexA and VP16 (Masuyama et al., 2012). The expression of the fusion protein is achieved via GAL4/UAS. Presence of calcium leads to dephosphorylation of LexA-VP16-NFAT by the calcium/calmodulin-dependent phosphatase calcineurin and subsequent translocation of the protein to the nucleus, where the LexA domain can drive expression of a reporter. In a similar approach, the TRIC system (transcriptional reporter of intracellular Ca²⁺) utilizes calcium-based binding of calmodulin (CaM) to its target peptide as a reporter of activity (Gao et al., 2015). CaM is fused to a transcriptional activation domain, while its target peptide is fused to a DNA binding domain of a transcription factor like GAL4. Binding of the fusion proteins then facilitates gene expression. The system acts on a slower timescale, reporting activity changes over long periods of time. Both the CaLexA and the TRIC system allow for calcium-based genetic access to active neurons and therefore modulation of those. Similar to optogenetic tools, where disruptions in neuronal activity have already been shown to alter the seizure phenotype under certain conditions (Giachello and Baines, 2015), these tools could be used to directly alter gene function during development.

Overall, Drosophila provides a versatile toolkit to induce, quantify and visualize seizure-like behavior. In combination with the vast genetic toolkit, large-scale screens to investigate the role of genes in neuronal function become possible.

Large-scale screens can be performed to study the effects of either genetic variance or compound treatment. Such genetic screens played an important role in forward-genetic approaches in the past, leading to the discovery of many seizure-related genes in Drosophila (Parker et al., 2011a). Investigating patient-derived variants of unknown significance on a similar scale is a goal, which becomes now feasible with the genetic toolkit of the fly. Such an approach has already led to the functional description of many epilepsy-related variants in single genes, for example by utilizing T2A-GAL4 technology and UAS-transgenes carrying either wild-type or gene variant for the gene OGDHL (Yoon et al., 2017; Yap et al., 2021). The number of epilepsy-related genes and variants will likely increase, and high-throughput screens are needed to meet the demand of functionally testing these variants. In a similar approach to the screen mentioned above, hundreds of variants regarding autism were functionally described in the fly (Marcogliese et al., 2022). Patient variants were expressed under endogenous promoters of fly orthologues or overexpressed in different tissues. These technologies were combined with behavioral monitoring and visualization of gene expression through fluorophores and provided a detailed analysis of different variants. Such a large-scale approach will likely be feasible for variants related to epilepsy as well. Combining such a screen with different methods of seizure induction or suppression might provide valuable information. The recent growth of epilepsy-associated genes also highlighted new molecular pathways and genetic mechanisms. Many of the latest findings have not been functionally explored in Drosophila yet (Takai et al., 2020). Investigating rare diseases has also led to a better understanding of common disease etiologies, e.g., in case of neurodegenerative diseases (Ma et al., 2022). This might also become the case for epilepsy caused by non-monogenic reasons.

Many substances have been evaluated for their potential to suppress seizures over the past decades (Loscher et al., 2020). Despite numerous newly marketed ASMs, the proportion of drug-resistant epilepsies is hardly changing (Chen et al., 2018). This unmet clinical need calls for innovative approaches in drug design and drug testing. Drosophila has been used for drug-screening before, based on bang-sensitive mutants, in which several anti-seizure drugs showed an ameliorating effect (Pandey and Nichols, 2011). In an RNAi-based approach, differences in expression patterns between wild-type, para^bss1 and PTX-fed flies uncovered several potential therapeutical targets (Lin et al., 2017). This allowed to identify and pharmaceutically influence the neuronal homeostasis gene pumilio, which was a rather unexpected target (Lin et al., 2018; Mulroe et al., 2022). Another unexpected discovery was that the DNA-topoisomerase I gene top1 also influences seizure-like phenotypes, emphasizing the potential of broad screens to identify novel therapeutic targets (Song et al., 2007). The rather unexacting approach of candidate compound screening in Drosophila also facilitates broader application and is less cost-intensive than in other models. It also allows for exploring less conventional treatments, e.g., agents used in traditional medicine (Dare et al., 2020).

Besides direct effects of drugs or genetic alterations, environmental effects should also be considered when screening for anti-seizure treatments. Though knock-in Drosophila models of GEFS+ and DS showed differences in behavior related to the variant severity in patients (Sun et al., 2012), this was not observed to the same degree in CRISPR-derived models for these syndromes, where the amino acid exchange took place at the same site in the protein (Roemmich et al., 2021). The striking differences seen in human variant carriers could not be replicated in flies. Additional genetic factors in patients or fundamental differences between humans and Drosophila could explain this observation. Indeed, salient phenotypic variations are not uncommon among individuals carrying identical variants (Helbig and Tayoun, 2016). Recent studies underline that common genetic variants are enriched in persons and families investigated for rare disease-causing variants, hinting at a broad array of influences, that shape a phenotype observed in a patient (Campbell et al., 2022; Oliver et al., 2022). Polygenic mechanisms that influence such epileptic phenotypes might be difficult to study in Drosophila, as they are large in numbers with low effect-sizes. Since the genetic background of the fly is very controllable though, decreasing background variation as much as possible might lead to a better understanding of the relationship between rare and common genetic variants. It has also been shown that, similar to possible therapeutic approaches in patients, a ketogenic-like diet can reduce seizure-like activity in several bang-sensitive mutants (Kass et al., 2016; Radlicz et al., 2019). In that regard, controlling living conditions also becomes important and is realized in Drosophila by standardizing fly husbandry techniques. In light of these environmental influences, epigenetic mechanisms should be considered as well. Epigenetics describes potentially heritable changes in gene expression that do not involve alterations in the DNA sequence and has become increasingly associated with epilepsy (Van Loo et al., 2022). As epigenetic mechanisms act on wide ranges in the genome, using a genetically well-defined and traceable organism like Drosophila might be beneficial to mitigate variance created by such effects.

Certain drawbacks should be considered when using Drosophila as a model for drug screening. The composition of fly food used for fly housekeeping or administration of drugs varies considerably between publications (Kruger and Denton, 2020). As not every compound is soluble in water, vehicles like DMSO may be used, which can have an additional effect on the fly. The amount of food consumed by the fly is also difficult to estimate. Colored additives can help with such assessments, as can radioactive tracers or platforms like FlyPAD, which allow for direct tracking of feeding behavior (Deshpande et al., 2014; Itskov et al., 2014; Shell et al., 2018). These methods inevitably increase the labor required to perform feeding experiments though, diminishing one of the advantages of using Drosophila. Metabolic processing of a compound and concentrations in different tissues of the fly are not assessed by such methods. The lack of an observable effect of a compound might be a false-negative finding due to these circumstances. Experiments should therefore be performed based on a standardized system. If an effect can be observed, further studies can be designed to solidify the finding.

Many of the mentioned technologies rely on tools for precise genome editing. CRISPR/Cas9 enables such a precise editing, and the technique has already transformed many areas of biological research, including Drosophila research (Port and Bullock, 2016). Through increased utilization of the technique, costs and time needed to generate flies with variants of interest will likely decrease more and more. The amount of patient variants with unclear disease relation rises as a product of advancements in next generation sequencing, and the need for functional studies of these variants rises as well (Bellen et al., 2019). The ability to model variants directly in the Drosophila genome is therefore highly desirable. CRISPR/Cas9 provides applications extending the precise genome editing though, such as targeted gene disruption (Port et al., 2020) or to produce UAS-transgene libraries based on human cDNA/ORF (Wei et al., 2020). Direct applications of CRISPR/Cas9-based gene editing in a clinical setting are also discussed, as the possibility to edit dysfunctional genes in somatic cells becomes especially interesting for genes with clear disease relationship (Carpenter and Lignani, 2021). A better understanding of functional aspects of these genes beforehand is critical though and the technology must overcome major obstacles like efficient delivery to targets and prevention of off-target effects, before it becomes clinically relevant.

Recently developed genetic techniques allow for increasingly precise control over gene expression. With CRISPR, a vast number of genes become accessible and can be hijacked or completely replaced to express GAL4 in their place instead (Kanca et al., 2022). For gaining temporal control over gene expression, optogenetic methods like PhotoGal4 or shineGal4 promise light-based on–off-switches, that circumvent the use of less precisely controllable factors like heat or uptake of a compound (de Mena and Rincon-Limas, 2020; di Pietro et al., 2021). Importantly, these methods build on established GAL4/UAS systems, which prevents already established lines from becoming obsolete. Directly controlling neuronal function as opposed to solely controlling gene function provides a valuable asset to the Drosophila toolkit as well, especially when investigating neuronal network function. With the ability to genetically target and control specific neuronal subtypes in Drosophila, optogenetic approaches promise control over neuronal function at any time (Kohsaka and Nose, 2021). As optogenetics can be used to induce seizure-like activity in Drosophila bang-sensitive mutants (Saras et al., 2017), the method will likely gain more traction, being better controllable and less invasive than electrostimulation. However, the established method of electrostimulation enables the assessment of seizure thresholds and can thereby detect subtle changes in seizure susceptibility. Optogenetics, on the other hand, has advantages in precisely targeting specific areas of the nervous system. The complementary use of both methods could help locate sites of seizure generation and determine seizure thresholds. Optogenetic approaches also provide additional ways to study network functions in flies during development (Giachello and Baines, 2015), or to also control gene expression (Chan et al., 2015; Lee et al., 2017). Optogenetics are furthermore researched in clinical settings, although many technical hurdles still have to be overcome before the technology is ready for concrete applications (Wykes et al., 2016).

Derived from the term optogenetics, the term chemogenetics offers similar techniques based on compound effects instead of light. It describes the design of proteins, which interact with small molecules that normally have no physiological relevance (Roth, 2016). In this broad category, Designer Receptors Exclusively Activated by Designer Drugs (DREADDs) gathered the most interest, as they allow for receptor activation through otherwise unrecognized substances. They are often G-protein coupled receptors and activate downstream targets upon activation. DREADDs are employed in epilepsy research and investigated for use in therapeutic settings, similar to optogenetic approaches (Mueller et al., 2022). A clinical application would also require a gene therapeutic approach. This raises concerns about the delivery of receptors, transport of designer agents across the blood–brain barrier and potential off-target effects (Krook-Magnuson and Soltesz, 2015). DREADDs have already been established in Drosophila (Becnel et al., 2013) and might provide another tool to alter neuronal functionality during developmental stage or in adult flies. As feeding assays for ASMs are established, the effectiveness of these approaches to suppress seizure-like behavior in flies could be compared. The methods could also be used in concert to evaluate the effects of ASMs while subsets of neurons are controlled through DREADDs or optogenetic approaches.

With the large number of techniques and genes available for studies in Drosophila, the amount of generated information becomes difficult to process. Utilizing large data resources and combining different resources will likely play an increasing role. The amount of data generated in terms of patient genomic data, variant effects, structural modeling, network and functional relationships and many more factors becomes ever increasing as well (Lhatoo et al., 2020). With the advent of single-cell RNA sequencing also in Drosophila, and extensive connectomes for the fly brain, even larger datasets will become available and combining data in an efficient manner becomes pivotal (Luo et al., 2018; Zheng et al., 2018; Scheffer et al., 2020; Li, 2021). Drosophila research provides and relies on large libraries of stocks and curated databases like FlyBase (Gramates et al., 2022), as well as tools to centralize and compare relevant information like MARRVEL (Wang et al., 2017a). Integration of these databases, as well as connectome and other network approaches will play a necessary role to allow for efficient communication in the Drosophila field and to make further progress in entangling epilepsy-related mechanisms (Wang et al., 2017a; Mehta et al., 2022).