The intricate mechanisms through which the brain and gut interact involve the release of endocrine mediators primarily derived from the hypothalamic-pituitary-adrenal (HPA) axis (Mayer et al., 2015). Part of the limbic system, the HPA axis is typically activated by stressors or systemic pro-inflammatory cytokines, and it primarily regulates emotional responses. Stress, in particular, induces emotional responses and stimulates hypothalamic activity. This stimulation triggers the secretion of corticotropin-releasing factor (CRF) (Stengel and Tache, 2010), which, in turn, stimulates the release of adrenocorticotropic hormone (ACTH) from the pituitary gland. ACTH then acts on the adrenal glands to promote the secretion of cortisol. Elevated levels of cortisol, a crucial stress hormone, interact with the gut to modulate various aspects of gastrointestinal function, visceral sensation, autonomic activity, and behavior (Ulrich-Lai and Herman, 2009).

At the intestinal level, these effector molecules exert effects on motility and secretion, intestinal permeability, local and systemic immune function and inflammation, as well as microbiota composition and function. For instance, increased intestinal permeability can disrupt the intestinal microbiota, resulting in elevated levels of pro-inflammatory cytokines (such as tumor necrosis factor α29) and the release of serotonin (5-HT) from enteroendocrine cells (Liebregts et al., 2007; Collins et al., 2012). These processes can enhance the sensitivity of visceral afferent nerves, amplifying stress-induced changes in motility, secretion, and permeability, which further heightens afferent signaling from the ENS (Keightley et al., 2015).

Both animal and human studies indicate that chronic or prolonged stress may lead to a sustained increase in the responsiveness of the brain’s central stress loop, ultimately resulting in functional and affective disturbances. Moreover, chronic or prolonged stress has been associated with the onset and exacerbation of symptoms in IBS (Whitehead et al., 1992). Patients with IBS often exhibit stress-induced alterations in gastrointestinal motility, gut sensation, autonomic regulation, and the response of the HPA axis (Whitehead et al., 2002). Over time, disruptions in the HPA axis can result in dysregulation of stress responses, pain modulation, and various other brain circuits, creating a maladaptive and mutually reinforcing cycle of brain-gut interactions.

Additionally, a diverse array of peptide hormones secreted by intestinal endocrine cells play a key role in regulating brain-gut signaling (Steinert et al., 2017). These intestinal peptides not only regulate gastrointestinal motility and sensitivity but also homeostatic brain circuits in the brainstem and hypothalamus through vagal afferent-terminal receptors. By doing so, they regulate food intake and energy homeostasis based on the body’s energy resources and nutritional needs. And then this regulatory neural network is involved in the motivational, emotional, and learning aspects of food reward (Alonso-Alonso et al., 2015). The combination of sensory signals from the gut, along with oral (taste) and extra-sensory inputs such as vision and odor, ultimately determines the rewarding value of food, thereby influencing appetite and feeding behavior (Weltens et al., 2014). Under normal circumstances, these gut peptides maintain a balance with the homeostatic network established by brain regions. However, when dysfunction occurs at any level of the brain-gut axis, this balance can be disrupted, leading to abnormal processing of signals, both physiological and noxious. Prolonged enhancement of afferent visceral signals may even impact mood, fear, anxiety, and other emotional processes (Berthoud and Morrison, 2008; Mayer et al., 2015).

Stress has the potential to trigger seizures, and the HPA axis is the primary regulatory system in the stress response (Cano-López and González-Bono, 2019). Different hormones have distinct effects on seizures: glucocorticoid levels are elevated in epilepsy patients, while deoxycorticosterone, an anticonvulsant drug, exhibits beneficial effects; Corticotropin-releasing hormone (CRH) and corticosterone promote the signaling of excitatory neurotransmitters like glutamate, which can induce seizures (Baram and Schultz, 1991; Reddy and Rogawski, 2002; Werner and Coveñas, 2017). Recent research has revealed a correlation between gut microbiota and the HPA axis. For instance, structural and functional changes in the gut microbiota due to stressful conditions regulate the expression of genes associated with the colonic CRH pathway (Vodička et al., 2018) (see Figure 2). Nevertheless, further investigation is required to understand the mechanism connecting gut microbiota, the HPA axis, and epilepsy.

Congruent lines of evidence have proposed that imbalances in neurotransmitters have been strongly linked to the development of epilepsy, with epileptic foci often exhibiting decreased levels of inhibitory neurotransmitters, such as GABA, and elevated levels of excitatory neurotransmitters like glutamate, dopamine, and norepinephrine (Singh and Goel, 2016). For instance, the disrupted balance of GABA and glutamate in brain tissue has been observed in patients with epilepsy (Fisher et al., 2017b). GABA is an inhibitory neurotransmitter widely distributed in the nervous system, and a decrease in GABA levels can lead to changes in neuronal cell permeability to synaptic interstitial chloride ions. This alteration can result in lower seizure thresholds and an increased susceptibility to epilepsy (Barrett et al., 2012; Peng et al., 2018). It has been found that various types of gut microbiome are capable of metabolizing and producing distinct neurotransmitters. For example, Enterococcus spp., Streptococcus spp., and Escherichia coli spp. can produce serotonin. Lactobacillus spp. and Bifidobacterium spp. are responsible for synthesizing GABA. Escherichia coli spp. and Bacillus spp., on the other hand, produce noradrenaline and dopamine (Tsavkelova et al., 2000; Özoğul, 2004; Shishov et al., 2009; Barrett et al., 2012; Özoğul et al., 2012; Strandwitz, 2018) (see Figure 2). Recent studies have shown the relative abundance of intestinal Ruminococcus and Coprococcus was positively correlated with brain tissue glutamate and glutamine levels (Sun et al., 2016; Zhang et al., 2021; Binh Tran et al., 2023). The gut microbiota, specifically the presence of beneficial bacteria such as A. mucinophilia and Parabacteroides, may provide protective benefits against seizures by modulating brain neurotransmitter levels, including GABA and glutamate in the hippocampus. Bacterial dysbiosis may alter GABA levels, leading to an exacerbation of seizures (Galland, 2014; McDonald et al., 2018). In addition, Barrett et al. investigated that Bifidobacterium spp., an important probiotic in the gut, not only stimulates inflammatory responses and boosts the immune system’s defenses against external pathogens but also secretes GABA that may traverse the blood-brain barrier (BBB) and enter the CNS. Notably, when mice were colonized with Ackermannia and Paramycetes in their gastrointestinal tracts, it led to reduced levels of excitatory amino acids (glutamate) in both the serum and intestines, along with increased expression of GABA in the hippocampus. This colonization exerted anti-seizure effects (Olson et al., 2018). However, current research has yet to confirm the direct entry of neurotransmitters produced by gut microbiota into the brain across the BBB and their role in shaping brain structure and function.

Previous studies have demonstrated that approximately 90% of 5-HT is produced and distributed in intestinal enterochromaffin cells (ECs), but the intestinal microbiota can play a role in its production and secretion process (Yano et al., 2015; Lu et al., 2021). The commensal microbiota has the ability to directly utilize luminal tryptophan to synthesize serotonin. Various bacteria including Lactococcus, Lactobacillus, Streptococcus, Escherichia coli, and Klebsiella have been found to produce serotonin by expressing tryptophan synthetase. Specifically, strains such as Lactococcus lactis subsp. Cremoris (MG 1363), Lactococcus lactis subsp. Lactis (IL 1403), Lactobacillus plantarum (FI8595), Streptococcus thermophilus (NCFB2392), Eschericchia coli K-12, Morganella morganii (NCIMB, 10466), Klebisella pneumoniae (NCIMB, 673), and Hafnia alvei (NCIMB, 11999) have been identified as capable of serotonin production. Furthermore, it has been observed that commensal bacteria, particularly spore-forming bacteria found in the mouse and human microbiota, can enhance serotonin biosynthesis in colonic ECs through a metabolite or cell component-dependent mechanism (Williams and Chryssanthopoulos, 1997; Sudo et al., 2004; O'Mahony et al., 2015; Yano et al., 2015; Clark and Mach, 2016; Eisenstein, 2016; Mandic et al., 2019). For instance, specific microbiota within the intestines of mice have been found to promote the production of 5-HT by regulating the rate-limiting enzyme responsible for its synthesis, tryptophan hydroxylase 1 (TPH1) (Yano et al., 2015). Depletion of 5-HT induced by reserpine reduces the threshold for seizures triggered by electrical stimulation in rats, thereby increasing their susceptibility to seizures (Wenger et al., 1973). Therefore, it is reasonable to hypothesize that the intestinal microbiota modulates the production and release of 5-HT from endocrine cells, subsequently affecting the electrical activity of the intestinal vagus nerve and the immune-inflammatory response. This modulation, in turn, leads to an increased susceptibility to epilepsy. Besides, studies have revealed a reduction in brain tissue N-acetylaspartate (NAA) levels in both epileptic patients and animal models of epilepsy (Ferrier et al., 2000; Gomes et al., 2007).

Specific intestinal microbiota, such as Firmicutes and Bacteroidetes, have the ability to ferment and break down insoluble dietary fiber, leading to the production of important metabolites known as short-chain fatty acids (SCFAs), including acetate, propionate, and butyrate (den Besten et al., 2013) (see Figure 2). These SCFAs play a significant role in promoting the maturation of microglia, and changes in microglial function and BBB permeability have been linked to seizure susceptibility (De Caro et al., 2019a). It is worth noting that in this context, patients with drug-resistant epilepsy exhibited significantly different levels of Firmicutes and Bacteroidetes phylum compared to those with drug-sensitive epilepsy and healthy controls, as demonstrated in studies by Xie et al. (2017), Peng et al. (2018), Lindefeldt et al. (2019), Lee et al. (2020), Şafak et al. (2020), and Lee et al. (2021). A rat model of post-traumatic epilepsy caused by lateral fluid percussion injury showed that the trauma itself did not result in significant alterations in the gut microbiota. However, the risk of developing post-traumatic seizure was strongly correlated with the disruption of gut microbiota and a decrease in intestinal SCFAs content prior to the traumatic event (Medel-Matus et al., 2022). In an absence seizure model using WAG/Rij rats, it was observed that brain tissue levels of SCFAs were significantly reduced, whereas the supplementation of butyrate effectively controlled their absence seizures. This can be attributed to the ability of butyrate to enhance mitochondrial function, protect brain tissue from oxidative stress, and prevent neuronal apoptosis, ultimately increasing the seizure threshold and reducing seizure intensity (Li et al., 2021a). As such, propionate supplementation has been shown to attenuate mitochondrial damage, hippocampal apoptosis, and neurological deficits, leading to a decrease in seizure intensity and an increase in seizure latency (Cheng et al., 2019). Furthermore, the KD has been utilized for nearly a century as a treatment for drug-resistant epilepsy in both children and adults (Xie et al., 2017; Olson et al., 2018; Zhang et al., 2018; Lindefeldt et al., 2019; Eor et al., 2021; Gong et al., 2021; Miljanovic and Potschka, 2021; Shearer et al., 2023). A high consumption of cruciferous and leafy vegetables, berries, and nuts on the KD has been linked to a positive impact on the profile of SCFAs produced by the gut microbiome (Gudan et al., 2022).

The gut microbiota plays a crucial role in maintaining the balance of the host immune system. Research conducted on mouse models of multiple sclerosis and stroke has highlighted the involvement of gut microbes in the regulation of autoimmunity, particularly impacting the development and function of CNS-resident immune cells, such as microglia M1 (Ochoa-Reparaz et al., 2010; Wang et al., 2014b; Benakis et al., 2016). For instance, when adult SPF mice were treated with antibiotics to remove bacteria, microglia reverted to an immature state. However, subsequent restoration to a normal state was possible through allogeneic gut microbiota transplantation, demonstrating the intimate connection between gut microbial signaling and the maintenance of a mature microglial state (Erny et al., 2015). Plus, disruptions in T cell subset homeostasis, including Th1, Th2, Th17, and Treg, due to intestinal malnutrition, can contribute to the development of autoimmune and inflammatory disorders (Lee and Kim, 2017).

Th1 cells, which secrete pro-inflammatory cytokines like IL-2, IL-12, TNF-alpha, and IFN-γ, play a critical role in promoting cellular immune responses (Luckheeram et al., 2012). Recent research has shown that alterations in the gut microbiota composition leads to the accumulation of specific amino acids, such as phenylalanine and isoleucine, which in turn promote the differentiation and proliferation of Th1 cells. Consequently, Th1 immune cells infiltrate the brain and engage in local communication with M1 microglia, causing a shift towards a pro-inflammatory state. This cascade ultimately contributes to neuroinflammation and cognitive dysfunction associated with conditions like AD (Wang et al., 2019).

IL-17, which is produced by Th17 and γδT cells during acute infections, has been shown to contribute to the pro-inflammatory response. Interestingly, in stroke models, there is an enrichment of γδT cells in the gut, which are then transported to the delicate membranes of the brain. This suggests a potential link between gut inflammation and the onset of stroke (Weaver et al., 2007). On the other hand, Treg cells play a crucial role in maintaining immune homeostasis by secreting the anti-inflammatory cytokine IL-10, which helps to suppress excessive immune responses (Zhu et al., 2010). Animal models of stroke have demonstrated that the deletion of Treg cells leads to a significant increase in the activation of resident and infiltrating inflammatory cells, including microglia and T cells. This heightened activation consequently attenuates the post-ischemic inflammatory response (Liesz et al., 2009). Moreover, Treg cells have been found to inhibit the differentiation of Th17 cells and the proliferation of γδT cells in the gut, thereby contributing to the maintenance of an anti-inflammatory environment (Huber et al., 2011).

Fecal calprotectin, a protein involved in intestinal inflammation, has emerged as a potential regulator of the inflammatory response in neurodegenerative diseases like AD and PD. A study showed that fecal calprotectin levels were elevated in nearly 70% of AD patients, suggesting its role in promoting neuroinflammation (Leblhuber et al., 2015). Similarly, elevated levels of fecal calreticulin in PD patients led to changes in the integrity of the intestinal epithelial barrier and the immune system (Mulak et al., 2017). Calreticulin is a calcium-binding protein consisting of the S100A8 and S100A9 heterodimer, which make up a significant proportion of the soluble protein content of neutrophils. Both S100A8 and S100A9 are capable of producing amyloid proteins and can form associations with other amyloids like β-amyloid (Aβ) and alpha-synuclein (α-syn), as well as amyloid oligomers and protofibrils (Wang et al., 2014a; Walsham and Sherwood, 2016). Macrophages and microglia secrete S100A9 during amyloid plaque formation, leading to its expression in neuronal cells. These effector molecules interact with Toll-like Receptor 4 (TLR4) and Receptor for Advanced Glycation End Products (RAGE) pathways, further activating microglia. The increased levels of calreticulin in cerebrospinal fluid (CSF) and the brain of AD patients may facilitate its co-aggregation with Aβ and amyloid formation (Wang et al., 2014a).

Among the intestinal microbiota amyloids, the curli protein (curli) produced by Escherichia coli is currently the most extensively studied. It shares similarities in primary and tertiary structures with CNS amyloids and forms biofilms with bacterial cells, providing resistance against physical and immunological factors (Cherny et al., 2005; Zhao et al., 2015). Exposure to intestinal bacterial amyloid can activate the immune system, thereby enhancing the response to endogenous amyloid production in the brain (Friedland and Chapman, 2017). It has been demonstrated that exposure of rats to bacteria capable of producing frizzled protein leads to increased α-syn deposition in neurons, as well as higher numbers of microglia and astrocytes in both the gut and brain. This exposure also results in elevated expression of Toll-like Receptor 2 (TLR2), interleukin-6 (II-6), and tumor necrosis factor (TNF) in the brain (Chen et al., 2016).

A cohort study conducted in northern Denmark, encompassing children born between 1998 and 2008, revealed a significant link between maternal infection during pregnancy and an increased risk of epilepsy in their offspring (Nørgaard et al., 2012). It is worth noting that this outcome is not specific to any particular pathogens or drugs but rather can be attributed to common processes associated with infections. Corroborating this finding, another cohort study examining Danish births from 1982 to 2012 reported a noteworthy 78% surge in the prevalence of epilepsy among children and young adults who had been hospitalized for an infection (Ahlers et al., 2019). The infants affected by group B streptococcus (GBS) infection, a leading cause of neonatal infections, faced a significantly elevated risk of developing epilepsy or other neurological disorders later in life (Yeo et al., 2019). The impact of infectious diseases on epilepsy is further amplified by epidemiological investigations from Norway, which demonstrated a significant rise in the incidence of infection-associated febrile seizures following the 2009 influenza A (H1N1) outbreak (Bakken et al., 2015). Furthermore, an association has been established between human herpesvirus type 6 (HHV-6) and medial temporal lobe sclerosis, a common cause of refractory epilepsy. For instance, infected individuals with medial temporal lobe sclerosis experience more frequent seizures and exhibit higher viral DNA loads indicative of HHV-6 involvement in the development of this condition through the induction of an aberrant immune-inflammatory response (Kawamura et al., 2015; Leibovitch and Jacobson, 2015). These findings collectively emphasize the critical role that infections (or immune-inflammatory response), both during pregnancy and throughout early life, play in shaping the risk and development of epilepsy.

Evidence from animal models further supports the notion that infections heighten the risk of developing epilepsy by engaging immunoinflammatory pathways, involving factors such as TNF-α, monocyte chemoattractant protein-1 (MCP-1), and others (Lv et al., 2014; Cusick et al., 2017; Patel et al., 2017; Sewal et al., 2017; Medel-Matus et al., 2018). For instance, studies utilizing rodent models have demonstrated that the injection of lipopolysaccharide, a component found in bacterial cell walls, triggers an upsurge in pro-inflammatory factors like TNF-α, IL-6, and IL-1β within brain tissue. This inflammatory response consequently lowers the threshold for seizures induced by chemical (PTZ) and electrical stimulation (Kołosowska et al., 2014; Russo et al., 2014; Leo et al., 2020; Hu et al., 2022) (see Figure 3). Intracerebral infection of C57BL/6J mice with the Daniels (DA) strain of Theiler’s murine encephalomyelitis virus (TMEV) leads to approximately 50% of the mice exhibiting acute behavioral seizures. The primary pathophysiological mechanism involves elevated levels of inflammatory factors in the hippocampus and limbic system, particularly IL-6 and TNF-α (Cusick et al., 2017). Remarkably, when IL-6 and TNF receptor 1 (TNFR1) and TNF receptor 2 (TNFR2) knockout mice were injected with TMEV, a significant reduction in seizures was observed, highlighting the key role of these inflammatory factors in the epileptic process (Cusick et al., 2017).

Th17 cells, identified as CD4⁺ T cell subset, hold a pivotal role in orchestrating the adaptive immune response (Weaver et al., 2007; Chow and Mazmanian, 2009; Huber et al., 2011; Lee and Kim, 2017). Their production of IL-17 can be modulated by specific intestinal flora, such as those belonging to the Bacteroidetes phylum. These gut bacteria can influence the activation and differentiation of Th17 cells, consequently altering the secretion of IL-17 (Ivanov et al., 2008) (see Figure 3). Notably, patients suffering from epilepsy exhibit markedly elevated levels of IL-17 in both cerebrospinal fluid and peripheral blood. In line with this, alterations in the expression of IL-17 showcase a positive correlation with both seizure frequency and severity. This intriguing link suggests that the gut flora may influence epilepsy susceptibility by intricately mediating the IL-17 pathway (Mao et al., 2013).

The pathogenesis of epilepsy is intertwined with the inflammatory pathways that are regulated by the brain-gut-microbiota axis (Gong et al., 2022b; Ding et al., 2022; Gernone et al., 2022; Mu et al., 2022). For instance, the composition of the gut flora may influence the immune function and modulate the susceptibility of hosts to seizures (Wu et al., 2016; Li et al., 2021b) (see Figure 3). Specifically, gut microbiota metabolizes dietary tryptophan into an aryl hydrocarbon receptor agonist. This agonist, in turn, interacts with microglia receptors, promoting their migration, apoptosis, and phagocytosis, ultimately leading to increased release of inflammatory factors (Ochoa-Reparaz et al., 2010; Wang et al., 2014a; Benakis et al., 2016; Rothhammer et al., 2018). Additionally, the activation of microglia further upregulates the expression of transforming growth factor-alpha (TGF-α) and vascular endothelial growth factor-B (VEGF-B) in astrocytes. As a consequence, an inflammatory response is triggered within astrocytes (Rothhammer et al., 2018). The crosstalk between astrocytes and microglia can potentiate an enhanced inflammatory response and disrupt the homeostasis of the BBB, facilitating the entry of peripheral blood immune cells and inflammatory factors into the CNS. This chronic inflammation, in turn, significantly elevates the risk of epilepsy (Rothhammer et al., 2018). Experimental evidence supports the role of gut flora in maintaining microglial immune function (Erny et al., 2015; Rothhammer et al., 2016). For instance, observations in mice reared in a sterile environment or treated with antibiotics to suppress gut flora reveal that microglia in these animals exhibit morphological, functional, differentiation, and activation defects, leading to innate immunodeficiencies against pathogens. Since there is no evidence to support that this pathological elimination of microbiomes does not affect the absorption of nutrients, exclusion of toxins, migration of cells, or other effects at a gut level, which could potentially have a multistep downstream effect on microglia. Therefore, it is reasonable to hypothesize that microglial changes are primarily an indirect result of changes in the microbiota. However, upon recolonization of the gastrointestinal tract, the immune function of these microglia is restored, underscoring the criticality of gut microbiota in regulating microglial function (Rothhammer et al., 2016). Moreover, abnormalities in the function of peripheral immune cells have been implicated in the development of epilepsy. T-lymphocytes and monocytes, upon migration to the CNS, can differentiate into macrophages and invade brain tissue, ultimately inducing seizures (Erny et al., 2015). Lin et al. (2021) demonstrated an increase in gut Klebsiella pneumoniae in patients with epilepsy. Subsequent animal studies have shed light on the impact of Klebsiella pneumoniae in the intestinal tract on seizure susceptibility and the activation of microglial cells to release inflammatory factors (Lin et al., 2021). This discovery highlights the intricate interplay between neuroinflammation and epilepsy pathogenesis. All told, the deep involvement of gut flora in both peripheral and neuroinflammation strongly suggests that gut flora-mediated neuroinflammation may play an important role in epileptogenesis that must not be overlooked (see Figure 3).

Intestinal microbiota exerts a regulatory role on the expression of key proteins involved in the integrity of the intestinal mucosal barrier, such as Occludin and Claudin. Animal studies have demonstrated a significant increase in BBB permeability in mice reared in a germ-free environment or treated with antibiotics to inhibit gut microbiota (Amasheh et al., 2011; Braniste et al., 2014). Dysbiosis of gut microbiota may downregulate the expression of Occludin and Claudin, leading to compromised integrity of the intestinal mucosal and BBB (see Figure 3). Consequently, this increased permeability facilitates the leakage of pathogenic microorganisms, metabolites, and toxins from the intestinal tract, and then the compromised barriers allow a large influx of immune cells and inflammatory factors to enter the CNS, ultimately triggering seizures (Rahman et al., 2018; Vezzani et al., 2019; Welcome, 2019). For instance, studies have found reduced expression levels of occludin and ZO-1, crucial components of the tight junctions in the BBB, in patients with drug-resistant temporal lobe epilepsy (Castañeda-Cabral et al., 2020). Besides that, the gut harbors a vast population of bacteria, many of which possess cell wall components, such as peptidoglycan (PGN) (Clarke et al., 2010). PGN and its fragments possess strong proinflammatory properties, signaling through Toll-like receptors (TLR), NOD-like receptors (NLR), and specialized PGN recognition proteins (such as PGLYRP1-4) (Laman et al., 2020). Alterations in the permeability of the intestinal mucosal barrier allow flora-associated metabolites, including PGN, to cross into the CNS. As a consequence, chronic inflammation is induced within the CNS, further contributing to the pathogenesis of epilepsy (Gales and Prayson, 2017). The collective findings from the aforementioned studies strongly implicate the involvement of intestinal flora in the regulation of epilepsy susceptibility through its influence on barrier permeability and immune-inflammatory responses. Although these studies provide valuable insights, it is important to note that only a limited number have investigated the intricate relationship between gut microbiota, immune-inflammatory responses, and epilepsy. Therefore, further well-designed studies are imperative to corroborate and expand upon these findings, in order to fully elucidate the role of the brain-gut-microbiota axis in the pathogenesis of epilepsy.

Significantly, epidemiological studies have established a clear association between inflammatory gastrointestinal diseases (GI) and the incidence of epilepsy. For instance, patients with epilepsy have been found to have up to eight times higher incidence of peptic ulcers compared to the general population (Keezer et al., 2016). Furthermore, a perforated peptic ulcer can trigger or complicate a generalized tonic-clonic seizure (Magon, 2010). Huang et al. (2010) demonstrated that even mild gastroenteritis could precede the development of benign infantile convulsions. Neurological complications are also observed in a considerable percentage of patients with inflammatory bowel disease (IBD), ranging from 0.25% to 47.50%. In severe cases of IBD, seizures of all types, including status epilepticus, can occur during the clinical course (Ferro and Oliveira Santos, 2021). Based on this evidence, it is plausible to hypothesize that some cases of epilepsy may be caused by autoimmune disorders that have manifestations in both the gut and the brain, with the direction of causality attributed to an extraneous factor (i.e., neither epilepsy nor the gut microbiome). Consider the example of Celiac disease, a well-documented systemic autoimmune condition characterized by gluten-triggered autoimmune intestinal villous atrophy, malabsorption, and a range of systemic and gastrointestinal symptoms. Approximately 10% of individuals with celiac disease experience neurological complications, including seizures. Conversely, about 0.78%–9.10% of epilepsy patients develop celiac disease (Djurić et al., 2012; Işıkay and Kocamaz, 2014). The precise mechanism behind these neurological manifestations remains poorly understood, likely linked to immune mechanisms. This hypothesis is supported by the presence of anti-Purkinje cells and anti-ganglioside antibodies in individuals with celiac disease who develop neurological symptoms (Pratesi et al., 2003). Moreover, drug-resistant epilepsy is more prevalent in children with celiac disease as a comorbidity. Remarkably, adherence to a gluten-free diet has led to the resolution of epilepsy in many celiac disease patients. Furthermore, maintaining a gluten-free diet alongside appropriate antiseizure medications has shown to reduce seizure frequency and severity in individuals with celiac disease and drug-resistant epilepsy (Swartwood et al., 2021). In these scenarios, it is plausible that a reduction in the inflammatory state, possibly through dietary changes, may significantly impact both gut microbiota and the severity of epilepsy. Nevertheless, these hypotheses necessitate further research to validate their potential implications.

Dr. Russell Wilder in the 1920s noted that seizures were halted during states of absolute fasting (Wheless, 2008; Winesett et al., 2015). With this observation, Dr. Wilder formulated a diet that emulated the effects of fasting by relying on a high-fat, moderate-protein, and low-carbohydrate composition. Specifically, the KD adheres to a calculated ratio of fat to the combined intake of proteins and carbohydrates, typically ranging between 2:1 and 4:1. To illustrate, a 4:1 ratio involves consuming four servings of fat for every one serving of proteins and carbohydrates. Therefore, 90% of the body’s required calories are derived from fat, while adequate protein is included to prevent the utilization of lean body mass. By metabolizing fat into ketone bodies, the brain is provided with an alternative fuel source to glucose, which is typically utilized when carbohydrates are consumed in excess.

Over the years, the dietary regimen developed by Dr. Wilder remains the foundation of KD therapy used today. Simultaneously, less strict variations of the original diet have also been introduced, including the Modified Atkins diet (MAD) and low glycemic index treatment (LGIT) (Kossoff and Doward, 2008; Pfeifer et al., 2008). Further, extensive exploration into the potential applications of the KD has expanded its use beyond epilepsy to other neurological disorders such as Alzheimer’s disease and Parkinson’s disease, among others. In recent years, the diet has even been investigated as a potential therapeutic intervention for non-neurological diseases. However, its foremost role remains as an effective treatment for epilepsy (Winesett et al., 2015; Xie et al., 2017; Olson et al., 2018; Zhang et al., 2018; Lindefeldt et al., 2019; Li et al., 2021b; Eor et al., 2021; Gong et al., 2021; Miljanovic and Potschka, 2021; Gudan et al., 2022; Shearer et al., 2023). In the pediatric age group, the KD has demonstrated significant benefits. In one study, it resulted in a 50% reduction in seizure frequency for at least 56% of patients, a 90% reduction in seizures for 32% of patients, and complete cessation of seizures in 16% of patients (Lefevre and Aronson, 2000). In a 2008 clinical trial conducted in England, children were randomly assigned either to receive the KD after a 1-month delay (treatment arm) or to undergo a 4-month delay (control arm) with no changes to their anticonvulsant medications. Five out of 54 patients in the treatment arm experienced a remarkable 90% reduction in seizures, while none of the patients in the control arm achieved such a response at the 4-month mark. Moreover, a greater than 50% decrease in seizures was observed in 38% of the patients in the treatment arm, compared to only 6% in the control arm. It is important to note that these findings were particularly significant considering the patients in the study were classified as quite refractory, having previously failed multiple antiseizure drugs (Neal et al., 2008).

Another study by Guzel et al. examined the effects of a 1-year KD intervention on 389 patients with drug-resistant epilepsy. The results revealed a significant improvement in the anti-seizure effect of the KD, with sustained efficacy over time. At 1, 3, 6, and 12 months, 65.8%, 74.7%, 70.6%, and 83.1% of the patients, respectively, demonstrated a positive response to the treatment (Guzel et al., 2019). The broad efficacy of the KD in treating various seizure types and childhood epilepsy syndromes, including symptomatic generalized, partial, and genetic epilepsies, is worth noting (Hartman, 2008; Hara et al., 2018). In fact, certain subgroups of epilepsy syndromes exhibit a particularly favorable response to the KD compared to available pharmacological treatments. One such remarkable subgroup is patients with glucose transporter (Glut-1) deficiency syndrome, where the efficacy of the KD may exceed 80% (Hartman, 2008; Leen et al., 2010). Recent research has delved into the effectiveness of combining the MAD with standard drug therapy (SDT) in the treatment of drug-resistant epilepsy. The findings of the study revealed that the combination of MAD and SDT was superior to SDT alone in reducing seizure frequency and improving overall outcomes, at the 6-month mark in adolescents and adults with non-surgical drug-resistant epilepsy (Manral et al., 2023). Moreover, the KD has demonstrated benefits beyond seizure control. It has been found to have a positive impact on the cognitive, emotional, verbal, and intellectual development of children (Zhu et al., 2016). However, KD or dietary treatment for seizures has thus far only shown effectiveness in pediatric populations. There is currently a lack of concrete evidence to support its benefits or effectiveness in adult populations. As a result, all the KD-related patients we discuss below are from studies focused on pediatric epilepsy.

A growing body of evidence suggests that the gut microbiota may play a significant role in the therapeutic effects of the KD for epilepsy (Olson et al., 2018; Spinelli and Blackford, 2018; Dahlin and Prast-Nielsen, 2019; Lindefeldt et al., 2019; Pittman, 2020; Ding et al., 2021; Yue et al., 2022; Özcan et al., 2022; García-Belenguer et al., 2023; Kundu et al., 2023; Shearer et al., 2023) (see Figure 4). A cross-sectional study conducted with 12 children suffering from drug-refractory epilepsy examined the impact of a 6-month KD treatment. The results were remarkable, with one patient experiencing complete cessation of seizures, three patients achieving a staggering 90%–99% reduction in seizure frequency, and five out of twelve children achieving a reduction of over 50% in seizure frequency. Cognitive and motor function improvements were observed in ten out of twelve children after 6 months on the KD. Importantly, a comparison of fecal samples taken before and after the 6-month KD intervention showcased that there was a notable decrease in the abundance of Bifidobacterium, Akkermansia, Enterococcaceae, and Actinomyces, along with a significant increase in the abundance of Subdoligranulum, Dialister, Alloprevotella, and Bifidobacterium (Gong et al., 2021). Another study conducted with 20 children suffering from refractory epilepsy who followed a KD for 6 months also yielded similar results. Two patients became completely seizure-free, three experienced a seizure reduction of 90% or more, five achieved a reduction between 50%–89%, and ten saw a reduction of less than 50%. Remarkably, all ten responders showed an improvement in their electroencephalogram (EEG) readings. In addition, analysis of fecal microbial profiles revealed a lower alpha diversity after KD therapy, along with significantly decreased abundance of Firmicutes and increased levels of Bacteroidetes (Zhang et al., 2018).

A significant increase in the abundance of the Bacteroidetes phylum within the intestinal flora was observed following KD treatment. This shift in microbial composition has been associated with the modulation of interleukin 6 and interleukin 17 secretion in dendritic cells, as well as promoting the production of SCFAs. These alterations play a key role in the reduction of seizure severity (Olson et al., 2018; Lindefeldt et al., 2019). Animal studies have also provided evidence supporting the efficacy of the KD in increasing seizure thresholds. In mice with a 6-Hz-induced refractory seizure model, the KD demonstrated the ability to raise seizure thresholds (Samala et al., 2008; Hartman et al., 2010; Olson et al., 2018). The KD also induced rapid and significant changes in the composition of the mouse gut microbiota. For instance, it led to an increase in the relative abundance of Akkermansia muciniphila and Parabacteroides (Samala et al., 2008; Hartman et al., 2010; Olson et al., 2018). To delve deeper into the relationship between the gut microbiota and the protective effects of the KD, Olson et al. conducted experiments involving germ-free feeding or antibiotic treatment (consisting of ampicillin, vancomycin, neomycin, and metronidazole) followed by 6 Hz electrical stimulation to induce epilepsy modeling. The results revealed that the protective effects of the KD were attenuated in germ-free fed and antibiotic-treated mice. However, when germ-free fed mice received the intestinal microbiota from normal mice, the seizure-protective efficacy was restored to that observed in normal mice receiving the KD. These findings strongly suggest that the presence of a healthy gut microbiota is necessary for the KD to exert its seizure-protective effects (Olson et al., 2018). Further investigations have indicated co-administration of Akkermansia muciniphila and Parabacteroides to the gastrointestinal tracts of germ-free or antibiotic-treated mice resulted in a notable increase in seizure thresholds upon 6 Hz electrical stimulation, supporting that the enriched intestinal flora associated with the KD possess anti-seizure properties (Frost et al., 2014; Vuong et al., 2017; Olson et al., 2018). Interestingly, colonization with only Akkermansia muciniphila or Parabacteroides alone did not restore the protective effect of the KD. This indicates that the presence of both bacteria is necessary for the anti-seizure effect of the KD (Olson et al., 2018). The metabolomic studies have suggested that administration of the KD resulted in decreased peripheral blood γ-D-glutamine levels, which was associated with an increase in the hippocampal GABA/glutamate ratio in mice, suggesting that the gut flora may modulate peripheral metabolites and central neurotransmitter metabolism, ultimately regulating seizure susceptibility (Frost et al., 2014; Vuong et al., 2017; Olson et al., 2018). The transplantation of KD-associated intestinal flora (Akkermansia muciniphila and Parabacteroides) into the gut of Kcna1 ^−/− sudden epileptic death model mice demonstrated elevated seizure thresholds, reduced seizure frequency, and shortened seizure duration. Instead, inhibition of both bacteria with antibiotics led to an increase in spontaneous tonic-clonic seizures in mice. These results highlight the broad applicability of the anti-seizure effects of the intestinal flora to various seizure types and model mice (Fenoglio-Simeone et al., 2009; Olson et al., 2018).

Remarkably, the manipulation of intestinal flora in models of IS using a KD and antibiotics has been shown to impact hippocampal mitochondrial function and subsequently alter epilepsy susceptibility. This observation suggests that targeting mitochondrial function-related flora could hold promise as a potential therapeutic strategy for epilepsy (Mu et al., 2022b). The major metabolite of ginsenosides Rb1, Rb2, and Rc in the intestinal microbiota, Ginsenoside compound K (GCK), has been found to have beneficial effects in a rat model of PTZ-induced epilepsy. GCK increases the levels of GABA, ultimately leading to a reduction in the intensity and prolongation of seizure latency (Zeng et al., 2018).

The disparities observed between animals in a germ-free environment and control animals are indeed fascinating (Fenoglio-Simeone et al., 2009; Frost et al., 2014; Vuong et al., 2017; Olson et al., 2018). However, it is crucial to approach the physiological effects of a germ-free environment with caution, as it represents an extreme condition that profoundly impacts the health and wellbeing of animals. Hence, it is imperative to build on these studies by conducting a thorough evaluation of the significance of the findings and delving deeper into the causation that precedes the observed factors.

FMT is an increasingly promising strategy for reconstructing the gut microbiota and has shown efficacy in various disease, such as epilepsy, Clostridium difficile infection, IBD, constipation, and other related disorders (Hartman et al., 2010; Borody and Khoruts, 2011; Surawicz et al., 2013; Reinisch, 2017; Olson et al., 2018). In a notable case study, a patient with both Crohn’s disease and refractory epilepsy underwent fecal transplantation, resulting in remarkable improvements. Following three consecutive transplants, the Crohn’s disease activity index (CDAI) significantly decreased from 361 to 131. Also, the patient’s seizures were completely controlled without the need for anti-seizure drugs 20 months after the fecal transplantation (He et al., 2017). However, this case is unique because it potentially involves an autoimmune etiology of Crohn’s disease as well as epilepsy. In another study, the administration of a probiotic mixture consisting of Lactobacillus acidophilus DSM32241, Lactobacillus plantarum DSM32244, Lactobacillus casei DSM32243, Lactobacillus helveticus DSM32242, Lactobacillus brevis DSM11988, Bifidobacterium lactis DSM32246, B. lactis DSM32247, and Streptococcus salivarius subsp. thermophilus DSM32245 in patients with drug-resistant epilepsy for four consecutive months yielded promising outcomes. Notably, 28.9% of the patients experienced a significant reduction of more than 50% in the frequency of seizures (Gómez-Eguílaz et al., 2018). Furthermore, the administration of probiotics, specifically Saccharomyces boulardii or Lactobacillus casei, to neonates within 24 h of birth has displayed promising effects in reducing the risk of seizures. This effect was observed regardless of whether the infants were infected with rotavirus or not. Conversely, rotavirus-infected infants who did not receive probiotics exhibited an increased risk of seizures (Yeom et al., 2019). One potential mechanism underlying the anti-seizure effect of Saccharomyces boulardii in rotavirus-infected infants involves the inhibition of rotavirus non-structural protein 4 (NSP4). NSP4 has been implicated in increasing reactive oxygen species (ROS) production, white matter damage, and immune-inflammatory responses. By inhibiting NSP4, Saccharomyces boulardii may mitigate these detrimental effects, thereby reducing the frequency and severity of seizures in these infants (Buccigrossi et al., 2014; Yeom et al., 2019) (see Figure 4).

Animal models have also provided evidence for the anti-seizure effects of probiotics (see Figure 4). Studies involving rat and mouse brain tissues have shown that the application of Lactobacillus rhamnosus and/or Bifidobacterium longum leads to increased expression of GABA and its receptor GABAR, resulting in a reduction in seizures (Bravo et al., 2011; Liang et al., 2017). Furthermore, the administration of probiotic supplements containing Lactobacillus rhamnosus, Lactobacillus reuteri, and Bifidobacterium infantis to PTZ-induced epilepsy model rat for 3 weeks demonstrated favorable outcomes. The intervention resulted in a decrease in epileptic activity levels, a significant reduction in seizure severity, and a partial improvement in spatial learning and memory in the epileptic rats. Importantly, the probiotic treatment also led to decreased concentrations of pro-inflammatory cytokines IL-6 and TNF-α, reduced levels of nitric oxide (NO) and malondialdehyde (MDA), increased total antioxidant capacity (TAC), and elevated levels of GABA in the brain tissue (Bagheri et al., 2019; Aygun et al., 2022a; Aygun et al., 2022b). Additional research utilizing animal models of epilepsy has corroborated the anti-seizure effects of probiotics. For instance, the supplementation of probiotics and NS (N-stearoyl-L-tyrosine) reduced population spike-long-term potentiation (PS-LTP) in kindled rats, thereby improving seizure-induced cognitive impairment (Gong et al., 2020). Other animal models have also demonstrated the potential anti-seizure effects of probiotics (Tahmasebi et al., 2020; Eor et al., 2021; Kilinc et al., 2021; Sabouri et al., 2021; Mu et al., 2022b; Wang et al., 2022; Ciltas et al., 2023).

Anti-seizure medications (ASMs) are typically the first line of treatment for controlling seizures in patients with epilepsy. However, a significant portion of these patients, estimated to be around 30%–40%, do not respond to at least two ASMs. Drug-resistant epilepsy patients also tend to exhibit dysregulation in their gastrointestinal ecological system, highlighting the potential involvement of the gut microbiota in epilepsy treatment (Peng et al., 2018). Drug-resistant patients have been found to possess higher alpha diversity, indicating a greater number of unique microbial types compared to drug-sensitive patients. Furthermore, specific microbiota has been associated with drug response. For instance, drug-sensitive patients tend to have a greater relative abundance of Bacteroides and Ruminococcus, while members of the Negativicutes group are overrepresented in drug-resistant epilepsy patients (Peng et al., 2018; Lee et al., 2021). Further evidence supporting the role of the gut microbiota in drug-resistant epilepsy comes from both animal models and pediatric cohort studies. These studies have demonstrated that the microbiota-based modulation of the anticonvulsant properties of the KD has shown promise in improving seizure control in these patients (Olson et al., 2018; Lindefeldt et al., 2019) (see Figure 4).

Previous research has demonstrated that intestinal inflammation can reduce the effectiveness of VPA, while therapy with SCFAs can effectively reduce inflammation, thereby enhancing the efficacy of VPA treatment (De Caro et al., 2019b). In animal models of epilepsy, the administration of lithium, valproate, and aripiprazole has been found to significantly increase the richness and diversity of microbial species. At the phylum level, valproate treatment has been shown to induce an increase in Actinobacteria and Firmicutes, while simultaneously decreasing the abundance of Bacteroidetes. At the genus level, several species including Clostridium, Peptoclostridium, Intestinibacter, and Christenellaceae were found to be increased following treatment with lithium, valproate, and aripiprazole compared to the control rats (Cussotto et al., 2019). Additionally, valproate treatment in rat models has been shown to decrease the relative abundance of S24-7 uncultbact and increase the relative abundance of Ruminococcaceae uncultured. Interestingly, VPA was also found to decrease the levels of two important SCFAs, namely, propionate and butyrate, in the cecum, while increasing the levels of isovalerate (Cussotto et al., 2019). Notably, mice exposed to sodium valproate demonstrated an increased ratio of Bacteroidetes to Firmicutes when compared to control mice (Sgritta et al., 2019). In clinical studies, it has been observed that the ratio of the phylum Firmicutes to Bacteriodetes is significantly higher in patients with new-onset epilepsy after 3 months of VPA treatment (Gong et al., 2022a). These findings provide further support for the role of gut microbiota composition in epilepsy and suggest that alterations in microbial populations may be associated with the effects of VPA therapy.

The anti-seizure drug lamotrigine and its ammonium salt complexes exhibit strong antibacterial activity against Gram-positive bacteria such as B. subtilis, S. aureus, and S. faecalis (Qian et al., 2009). Clonazepam, another anti-seizure drug, is metabolized by intestinal flora, which can potentially increase the drug’s toxic effects (Zimmermann et al., 2019). Then carbamazepine has been found to induce specific phenotypes in gut microbiota (Watkins et al., 2017; Gong et al., 2022; Ilhan et al., 2022). In a study involving epilepsy patients, after 3 months of carbamazepine treatment, the frequency of seizures was reduced by more than 50%. The abundance of Actinobacteria at the phylum level, which was initially higher in epileptic patients compared to healthy controls, decreased significantly after treatment. At the genus level, the abundances of bacteria such as Escherichia/Shigella, Streptococcus, Collinsella, and Megamonas were significantly higher in epilepsy patients before treatment but decreased significantly after treatment (Gong et al., 2022).

ASMs typically possess a narrow therapeutic window and are associated with numerous adverse effects, particularly impacting the gastrointestinal tract. Between 10% and 30% of epilepsy patients experienced intolerable side effects from these medications, leading them to discontinue treatment, particularly when undergoing polytherapy (Wiebe et al., 2008). Many ASMs have a direct impact on the enteric nervous system, influencing gut motility, which is evident in the gastrointestinal adverse effects of these medications (Jahromi et al., 2011; Minton et al., 2011; Al-Beltagi, 2021; Al-Beltagi and Saeed, 2022; Nevitt et al., 2022). For example, carbamazepine can lead to a range of gastrointestinal adverse effects, including dry mouth, mouth sores, glossitis, loss of appetite, dysphagia, nausea, vomiting, heartburn, gastritis, stomach/abdominal pain, constipation, and diarrhea (Anttila and Valtonen, 1992). Similarly, ethosuximide can induce anorexia, nausea, vomiting, gastric pain, diarrhea, and gastric and intestinal atony with decreased peristaltic activity (Kristev et al., 1994; Zagorchev et al., 1998). Phenobarbital can result in sore throat, diarrhea, swelling of the tongue/throat, nausea, vomiting, constipation, dysphagia, and heartburn (Jahromi et al., 2011; Al-Beltagi and Saeed, 2022). Phenytoin can cause changes in taste sensation, gingival overgrowth, sore throat, mouth ulcers, diarrhea, nausea, vomiting, constipation, dysphagia, heartburn, and reduced gastrointestinal absorption of calcium (Bell et al., 1979). Valproate also leads to gastrointestinal adverse effects, including diarrhea, nausea, vomiting, constipation, dysphagia, and gastritis with heartburn (Abdullah and Mousheer, 2020). This influence on the gut may significantly contribute to alterations in the gut microbiome. However, despite our comprehensive review of the existing research literature, we have found a lack of studies specifically examining the microbiota alterations induced by the gastrointestinal adverse effects of ASMs.

Carbamazepine, lamotrigine, and topiramate have been found to inhibit the growth of more than ten strains of bacteria when combined with the syrup excipient propyl-paraben. Ilhan et al. (2022) studied the effects of sweeteners and benzoates commonly found in ASM syrup. They found that the impact of these chemicals was concentration-dependent, with parabens and sweeteners showing increased toxicity or proliferation with higher concentrations. Bifidobacterium species were negatively affected by high concentrations of propyl-paraben, while methyl-paraben did not have the same effect. These results align with previous research by Crovetto et al., who reported a greater toxic effect of propyl- and butyl-parabens compared to methyl-paraben on E. coli and Staphylococcus aureus under aerobic conditions (Cussotto et al., 2019). Interestingly, the presence of various artificial sweeteners in ASMs formulations stimulated the growth of certain gut bacterial strains. The active ingredients that exhibited greater toxicity towards bacterial strains also displayed toxicity towards HT-29 cells (Watkins et al., 2017; Ilhan et al., 2022). Notably, previous studies have suggested that common food and drug additives, including artificial sweeteners, can alter the composition and function of the gut microbiota in batch-culture systems (Gerasimidis et al., 2020). Aspartame and benzoate have been shown to increase the relative abundance of Bifidobacterium in these mixed cultures, indicating a potential prebiotic effect (Gerasimidis et al., 2020). In addition, the supernatant of Bifidobacterium longum was found to reduce the cytotoxic effects of carbamazepine and lamotrigine. Similarly, the supernatants of Akkermansia muciniphila or mixed bacterial communities reduced the expression of drug resistance genes in HT-29 cell lines (Ilhan et al., 2022). In conclusion, these findings reveal that medications targeted for humans may inadvertently affect the gut microbiome, potentially exerting anti-commensal effects. This highlights the complex relationship between human-targeted medications and the gut microbiome, emphasizing the importance of understanding and modulating the microbiome to optimize treatment outcomes.

In general, the gut microbiota appears to play a significant role in the functioning and development of fundamental physiological processes, and it may also have an impact on central neural processes through the microbiota-gut-brain (MGB) axis. Both preclinical and clinical studies suggest that the microbiome could potentially modulate seizures and contribute to the pathogenesis of epilepsy. Various interventions, such as changes in diet, supplementation, and medication, have the potential to directly and indirectly affect the MGB axis. Researching the effects of these interventions could lead to a better understanding of epilepsy, the identification of biomarkers, and the development of new therapeutic options. That’s where our interest in the field lies. However, investigating the MGB axis and the role of gut supplementation in epilepsy is challenging due to the numerous potential pathways and variables involved. To date, only a few limited studies have been conducted, making it premature to draw conclusions. We also acknowledge that it is essential to conduct studies with the same rigor as pharmaceutical drug development trials, including taxa and metabolomic analyses using standard methodologies.