Endocannabinoid-Mediated Control of Neural Circuit Excitability and Epileptic Seizures

Endocannabinoid Signaling Mediated by CB₁ Receptor

One of the most important discoveries in the field of cannabinoid research is the discovery that endocannabinoids function as a retrograde messenger at synapses—endocannabinoids mediate signals from depolarized postsynaptic neurons to presynaptic CB₁ receptors, leading to a transient suppression of synaptic transmission (Kreitzer and Regehr, 2001; Ohno-Shosaku et al., 2001; Wilson and Nicoll, 2001). Furthermore, activation of G_q/11 coupled-receptors in neurons releases endocannabinoids that retrogradely act on presynaptic CB₁ receptors and induce transient synaptic suppression (Maejima et al., 2001; Varma et al., 2001). Soon after these discoveries, endocannabinoid-mediated long-term suppression of synaptic transmission was reported (Gerdeman et al., 2002; Marsicano et al., 2002; Robbe et al., 2002). Majority of short-term and long-term depression (STD and LTD, respectively) of synaptic transmission through the CB₁ receptor is mediated by 2-AG (Kano et al., 2009; Luchicchi and Pistis, 2012). On the other hand, AEA is reported to be involved in LTD at certain types of synapses (Chávez et al., 2010; Khlaifia et al., 2013; Mathur et al., 2013). We will first describe CB₁ receptor-dependent STD that is mediated by 2-AG (Figure 1), and then discuss the mechanisms of LTD that is mediated by 2-AG or AEA.

CB₁ receptor-dependent STD occurs when a postsynaptic neuron is strongly depolarized and the intracellular Ca²⁺ level increases following Ca²⁺ influx through voltage-gated calcium channels. This STD of excitatory or inhibitory synaptic transmission is called depolarization-induced suppression of excitation (DSE) or inhibition (DSI), respectively. DSE/DSI was found to be totally abolished in mice deficient in diacylglycerol lipase α (DGLα), the major 2-AG producing enzyme from diacylglycerol (Bisogno et al., 2003), demonstrating that 2-AG is responsible for endocannabinoid-mediated STD (Gao et al., 2010; Tanimura et al., 2010). Biosynthesis of 2-AG induced by a large increase in the intracellular Ca²⁺ level alone is termed Ca²⁺-driven endocannabinoid release (Ca-ER) (Maejima et al., 2005; Hashimotodani et al., 2007a; Ohno-Shosaku et al., 2007). Another pathway for 2-AG production is initiated by the activation of G_q/11 protein-coupled receptors, such as group I metabotropic glutamate receptors and M₁/M₃ muscarinic acetylcholine receptors, which is termed receptor-driven endocannabinoid release (RER) (Maejima et al., 2001; Fukudome et al., 2004; Hashimotodani et al., 2007a). Importantly, simultaneous weak activation of G_q/11 protein-coupled receptors and small elevation of intracellular Ca²⁺ concentration effectively induces the production of 2-AG much larger than the 2-AG level by either stimulation alone, which is termed Ca²⁺-assisted receptor-driven endocannabinoid release (Ca-RER) (Hashimotodani et al., 2005, 2007a; Maejima et al., 2005). In postsynaptic neurons, the amount of 2-AG used for retrograde messengers might be regulated by the 2-AG-hydrolyzing enzymes α/β-hydrolase domain (ABHD) 6 and 12 (Blankman et al., 2007) and by a 2-AG oxidizing enzyme, cyclooxygenase (COX)-2 (Kozak et al., 2000). Due to its lipophilic nature, 2-AG is thought to pass through the plasma membrane, although it should cross the hydrophilic synaptic cleft toward presynaptic terminals; the details of these processes remain unknown.

After crossing the synaptic cleft, 2-AG binds to and activates CB₁ receptors at the presynaptic terminal as a full agonist (Hillard, 2000). Activation of the CB₁ receptor suppresses neurotransmitter release mainly by inhibiting voltage-gated calcium channels (Pertwee, 1997; Huang et al., 2001), activating G protein-gated inwardly rectifying potassium channel (Henry and Chavkin, 1995) and by suppressing the activity of the presynaptic protein Munc18-1 through the phosphorylation of extracellular-regulated kinase (Schmitz et al., 2016). The action of 2-AG is terminated when it is degraded by monoacylglycerol lipase (MGL) (Hashimotodani et al., 2007b; Zhong et al., 2011; Tanimura et al., 2012) or potentially by presynaptic COX-2 (Wang et al., 2005).

In addition to STD, presynaptic forms of 2-AG-mediated LTD have been reported in several brain regions. At inhibitory synapses in the hippocampus, the mechanisms of 2-AG production and release in LTD are similar to those in STD, but LTD induction requires 5 to 10 min of continuous CB₁ receptor activation and presynaptic activity (Chevaleyre and Castillo, 2003). Activation of the CB₁ receptor mobilizes G_{α i/o} protein and decreases cAMP production and protein kinase A (PKA) signaling, which eventually results in a persistent reduction in the efficacy of neurotransmitter release (Chevaleyre et al., 2007). It is noteworthy that the LTD is sustained without further activation of CB₁ receptors after induction. Protein synthesis (Younts et al., 2016) as well as presynaptic proteins including RIM1α (Chevaleyre et al., 2007), calcineurin (Heifets et al., 2008), and potassium channels (Yasuda et al., 2008) are required for the induction of CB₁-mediated LTD at inhibitory synapses in the hippocampus.

As described previously, the other major endocannabinoid AEA is crucial for the induction of CB₁-mediated LTD at inhibitory synapses in the striatum (Mathur et al., 2013) and amygdala (Azad et al., 2004). Moreover, AEA-mediated LTD through TRPV1 was observed at perforant path–granule cell synapses in the dentate gyrus (Chávez et al., 2010) and in D₂ dopamine receptor-positive medium spiny neurons in the nucleus accumbens (Grueter et al., 2010). Although the biochemical pathways for AEA synthesis are not fully understood, the cascade leading to the production of AEA involves the synthesis of N-acylphosphatidylethanolamine (NAPE) from phospholipids by N-acyltransferase (Cadas et al., 1997) and the degradation of NAPE by NAPE-specific phospholipase D to AEA (Okamoto et al., 2004; Leishman et al., 2016). AEA is cleaved to arachidonic acid by fatty acid amide hydrolase (FAAH) (Cravatt et al., 1996). AEA is also degraded by COX-2 in postsynaptic neurons (Yu et al., 1997).

In addition to the activity-dependent phasic suppression of synaptic transmission by 2-AG, endocannabinoid signaling through the CB₁ receptor also mediates tonic suppression of synaptic transmission at inhibitory synapses in the CA3 (Losonczy et al., 2004) and CA1 (Neu et al., 2007; Lee et al., 2015) and at mossy cell–dentate granule cell excitatory synapses in the dentate gyrus (Jensen et al., 2021). The tonic CB₁ signaling could be due to the continuous release of endocannabinoids (Neu et al., 2007; Lee et al., 2015) or constitutively active CB₁ receptors (Lee et al., 2015; Jensen et al., 2021).

Detailed mechanisms of endocannabinoid-mediated regulation of synaptic transmission through the CB₁ receptor have been described in several excellent reviews (Heifets and Castillo, 2009; Kano et al., 2009; Castillo et al., 2012).

CB₂ Receptor-Mediated Signaling

The CB₂ receptor is the other canonical G_i/o-coupled cannabinoid receptor that mediates endocannabinoid signaling (Munro et al., 1993). In the central nervous system, the expression of CB₂ receptor mRNA in neurons has been confirmed using fluorescent in situ hybridization (Li and Kim, 2015; Stempel et al., 2016). Unfortunately, the CB₂ receptor protein expression in neurons remains elusive because of the lack of specific CB₂ receptor antibodies. However, recent studies using CB₂ knockout mice demonstrated that CB₂ receptors contribute to the regulation of neuronal excitability and synaptic transmission. CB₂ receptor signaling is required for maintaining excitatory synaptic transmission and LTP at Schaffer collateral–CA1 pyramidal cell synapses (Li and Kim, 2016). In contrast, CB₂ receptor signaling has a long-lasting hyperpolarizing effect in hippocampal CA3 pyramidal cells (Stempel et al., 2016). Long-lasting hyperpolarization mediated by CB₂ receptor signaling has also been reported in layer 2/3 regular spiking non-pyramidal cells in the somatosensory cortex (Stumpf et al., 2018). Furthermore, firing of dopamine neurons in the ventral tegmental area has been reported to decrease through activation of CB₂ receptors in vivo (Zhang et al., 2014). Thus, the effects of CB₂ receptor signaling on neuronal network activity differ depending on brain areas and cell types.

Involvement of 2-AG and AEA

When the concentration of 2-AG was measured in rat hippocampi 15 min after the initiation of pilocarpine-induced SE, an approximately 1.5-fold increase in 2-AG levels compared to vehicle-treated controls was observed (Wallace et al., 2003). Similarly, the 2-AG level in brain tissues 100 min after kainate treatment in 8-week-old rats was 1.5-fold higher than that in vehicle-treated controls (Fezza et al., 2014). Recently, the endocannabinoid sensor GRABeCB2.0 has been developed (Dong et al., 2021). By imaging fluorescent signals representing released endocannabinoids by using GRABeCB2.0, a transient surge of endocannabinoid release was successfully detected during afterdischarges evoked by the electrical stimulation of the contralateral hippocampus (Farrell et al., 2021). Importantly, the authors’ pharmacological examination demonstrated clearly that 2-AG, but not AEA, is produced during afterdischarges.

However, in sharp contrast to the above results, Marsicano et al. (2003) reported that the concentration of 2-AG in the hippocampi of adult mice 20 min after kainate injection was not different from the pre-treatment level. Instead, the concentration of AEA increased significantly to up to 300% of the pre-treatment level and decreased back to baseline within 1 h after kainate treatment (Marsicano et al., 2003). These results are consistent with a recent study showing no change in the levels of 2-AG and AEA in the cerebral cortex, striatum, hippocampus, thalamus, hypothalamus, and cerebellum 1 h after kainate injection (Lerner et al., 2017). In another study using the kindling model, there were no changes in 2-AG and AEA levels in the hippocampus 4 weeks after PTZ kindling (Hansen et al., 2009).

The reason for these discrepant results of 2-AG and AEA levels could be the difference in the timing of sampling of brain tissues after SE or kindling epileptogenesis. As observed in the endocannabinoid imaging using GRABeCB2.0, the endocannabinoid level changes rapidly depending on neural activity, which is ascribed to rapid degradation of endocannabinoids (Farrell et al., 2021). In addition, SE occurs intermittently after kainate or pilocarpine injection. Therefore, endocannabinoid levels could differ significantly depending on the timing of sampling of brain tissues. It is possible that elevation of endocannabinoid concentration after seizures was not detected if sampling of brain tissues was performed during inter-ictal periods or after mild seizures.

Changes in the expression of enzymes related to endocannabinoid synthesis and degradation have also been investigated (Figure 2). We investigated the expression of DGLα in the hippocampal CA1 and dentate gyrus of mice that received unilateral intrahippocampal kainate injection and reported a significant decrease in the expression of DGLα 2 weeks after kainate injection (Sugaya et al., 2016). The decreased DGLα expression in the hippocampi of epileptic mice is consistent with that in the hippocampi of patients with mTLE (Ludányi et al., 2008). In contrast, the level of MGL mRNA was unchanged 1 h and 7 days after kainate-induced SE (Terrone et al., 2018). These results suggest that production of 2-AG is decreased in epileptic brains. Moreover, decreased expression of DGLα may affect the process of epileptogenesis (Table 1). We demonstrated that 2-AG prevented epileptogenesis in a perforant path kindling model, and that loss of 2-AG promoted the kindling process (Sugaya et al., 2016). These data are consistent with the results obtained in amygdala-based (von Rüden et al., 2015a) and corneal (Griebel et al., 2015) kindling models. The 2-AG-mediated anti-epileptogenic effect could be due to the preventive effect of 2-AG against neuronal injury. Preventive effects of MGL inhibitors against neuronal degeneration in vitro (Kallendrusch et al., 2012) and in a kainate-induced SE model in vivo (Terrone et al., 2018) have also been reported. The MGL inhibitor CPD-4645 also decreased the mRNA level of the inflammation marker interleukin-1β at 1 h and at 7 days after kainate-induced SE (Terrone et al., 2018).

TABLE 1

Epileptogenesis and ictogenesis modulated by manipulations of endocannabinoid signaling.

Open in a separate window

Summary of the effects of enhancing or suppressing endocannabinoid signaling (2-AG, AEA, CB₁ or CB₂) on epileptogenesis and ictogenesis in various rodent seizure models.

Cells in the leftmost column indicate the manipulations of endocannabinoid signaling molecules and those in the first and second rows show rodent seizure models.

The observed effect in each study was represented as follows; blue, suppression of epileptogenesis or ictogenesis; black, no change; red, promotion of epileptogenesis or ictogenesis; ?, unknown.

Excitatory, results observed in mice with excitatory neuron specific deletion of CB₁ receptors; Inhibitory, results observed in mice with inhibitory neuron specific deletion of CB₁ receptor.

However, in sharp contrast to the results mentioned above, DGLα blockade by RHC80267 for 7 days initiated immediately after the termination of pilocarpine-induced SE decreased cell death and the frequency of spontaneous seizures (Ma et al., 2014), indicating that 2-AG promotes SE-induced epileptogenesis. The pro-epileptogenic effect of 2-AG during the initial 7 days after SE may be explained by the downregulation of CB₁ receptor signaling. Considering the fact that 5 days of 2-AG upregulation by treatment with the MGL inhibitor JZL 184 [16 mg/kg, intraperitoneal (i.p.)] induced the downregulation of CB₁ receptor signaling in non-epileptic wild-type mice (Kinsey et al., 2013), it is likely that downregulation of CB₁ receptor signaling at excitatory synapses occurs after sustained increase in 2-AG levels after SE. Another potential explanation for the pro-epileptogenic effects of 2-AG would be prostaglandin-mediated stroke-like hypoperfusion events that occur after seizures (Farrell et al., 2021). The hydrolysis of 2-AG by MGL produces arachidonic acid which is broken down to vasoactive prostaglandins by COX-2. Therefore, surge of 2-AG production during seizures results in the vasoconstriction and hypoxia around the seizure focus for tens of minutes (Farrell et al., 2021), which could potentially damage the tissues around the seizure focus and promote epileptogenesis. Taken together, it is thought that the 2-AG level initially increases during SE and then decreases along with the decreased expression of DGLα in the chronic phase of epileptogenesis. The role of 2-AG in epileptogenesis may differ depending on the level and the duration of 2-AG surge.

In contrast to the aforementioned effects of 2-AG, increase in the level of AEA by treatment with the FAAH inhibitor URB597 did not affect kindling epileptogenesis (Wendt et al., 2011). As for SE-induced epileptogenesis, elevation of the AEA level is reported to exert different effects depending on the degree of elevation (Table 1)—a mild increase prevented (Mikheeva et al., 2017) whereas a strong increase promoted cell death (Clement et al., 2003).

Involvement of 2-AG

The role of 2-AG signaling in ictogenesis has been investigated using DGLα knockout mice (Sugaya et al., 2016). After kainate injection (30 mg/kg, i.p.), DGLα knockout mice showed shorter latency to tonic-clonic seizures and a higher mortality rate than wild-type mice. In addition, DGLα knockout mice exhibited longer afterdischarges in the dentate gyrus than wild-type mice when the perforant path was electrically stimulated. Consistently, MGL inhibitor, JZL184, ameliorated kainate-induced SE (Fezza et al., 2014; Sugaya et al., 2016). These results suggest that 2-AG is crucial for the suppression of seizures in the hippocampus.

A contrasting result was reported in the excitatory neurons of the dentate hilus, CA1, and CA3 of MGL-overexpressing mice (Guggenhuber et al., 2015). In pyramidal neurons of these mice, DSE was abolished, whereas DSI was intact, possibly because of the immediate degradation of 2-AG at excitatory synapses. However, no differences were observed in the severity of kainate-induced seizures (35 mg/kg, i.p.) compared to those in control mice, which is contrast to aforementioned results in DGLα knockout mice (Sugaya et al., 2016). The difference in seizure susceptibility between these two mouse lines could be ascribed to the difference in the concentration of 2-AG in the hippocampal tissues. In the hippocampi of MGL-overexpressing mice, the level of 2-AG was approximately half that in control mice, while that in DGLα knockout mice was reduced to approximately 1/10 of that in wild-type mice (Tanimura et al., 2010). Furthermore, DSE requires a much higher 2-AG level than DSI (Ohno-Shosaku et al., 2002). Therefore, even in the absence of DSE in the hippocampus of MGL-overexpressing mice, the occurrence of kainate-induced seizures might be suppressed by the remaining level of 2-AG. Another possibility would be that the kainate-induced seizures, which lasted for more than 3 h in the study, resulted in 2-AG production while a few seconds of depolarization failed to produce sufficient 2-AG for DSE. This prolonged 2-AG synthesis might have overwhelmed the capacity of 2-AG degradation by the overexpressed MGL.

As expected, based on the crucial role of 2-AG in the prevention of kainate-induced seizures, augmentation of 2-AG concentration over the physiological level has been reported to ameliorate seizures (Table 1). Administration of an MGL inhibitor to wild-type mice reduced the severity of various types of evoked seizures (Griebel et al., 2015; von Rüden et al., 2015a; Sugaya et al., 2016). Moreover, a recent study using the MGL inhibitor CPD-4645 demonstrated that the ameliorating effect of MGL inhibition on kainate-induced SE involved 2-AG-mediated activation of the CB₁ receptor initially, but later was predominantly mediated through mechanisms independent of CB₁ receptor signaling (Terrone et al., 2018), which is consistent with the potential role of the CB₂ receptor in seizure protection.

WWL123, an inhibitor of the postsynaptic 2-AG degrading enzyme ABHD6, is also capable of reducing the severity of PTZ-induced seizures (Naydenov et al., 2014). Surprisingly, the suppressive effect of WWL123 on PTZ seizures was observed in CB₁ and CB₂ double knockout mice and therefore is thought to be independent of the CB₁ and CB₂ receptors. The primary target of increased 2-AG at postsynaptic neurons could be GABA_A receptors, as 2-AG directly activated them (Naydenov et al., 2014). In contrast, the MGL inhibitor SAR127303 did not change the minimal dose of PTZ required to induce seizures (Griebel et al., 2015), suggesting that postsynaptic increase in 2-AG is important.

These results demonstrate the suppressive effect of 2-AG on seizures in the non-epileptic brain (Table 1). In parallel, the role of 2-AG signaling in spontaneous seizures in the epileptic brain has also been investigated (Table 1). We have reported that the MGL inhibitor JZL184 has a suppressive effect on the frequency of spontaneous seizures after kainate-induced SE (Sugaya et al., 2016). In genetically epileptic mice, pharmacological blockade of ABHD6 by means of WWL123 completely suppressed the occurrence of spontaneous seizures in the R6/2 mouse strain (Naydenov et al., 2014). These results suggest that augmentation of 2-AG signaling strongly suppresses spontaneous seizures in the epileptic brain (Figure 2 and Table 1).

Involvement of AEA

Direct application of AEA or augmentation of AEA levels by blocking its degradation using an FAAH inhibitor suppresses the seizures induced by transcorneal electrical stimulation (Wallace et al., 2002). Since such a seizure-suppressive effect was abolished by co-administration of the CB₁ antagonist SR141716A, the suppressive effect of AEA on evoked seizures was thought to be mediated by the CB₁ receptor. FAAH inhibitors also suppress seizures induced by kainate (Karanian et al., 2007; Fezza et al., 2014) and PTZ (Vilela et al., 2013; Table 1).

However, contradictory results were obtained in a study using FAAH knockout mice. AEA levels in the cortex, hippocampus, and cerebellum were 10 times higher in FAAH knockout mice than in wild-type mice, presumably because of the slower degradation of AEA. However, kainate-induced seizures were more severe in FAAH knockout mice than in wild-type mice (Clement et al., 2003). Moreover, administration of AEA in FAAH knockout mice further aggravated the kainate-induced seizures and cell death. A CB₁ blocker antagonized the pro-convulsive effect of AEA, suggesting that the potential mechanism underlying the effect of AEA depends on the CB₁ receptor. It is possible that AEA aggravates seizures by suppressing GABA release from inhibitory presynaptic terminals (Lourenço et al., 2011).

As described above, AEA acts as a full agonist of TRPV1 (Smart et al., 2000). TRPV1 is a Ca²⁺-permeable cation channel, and its activation depolarizes neurons. TRPV1 activation is reported to suppress 2-AG synthesis and increase tonic inhibition through the reduction of tonic 2-AG signaling (Lee et al., 2015). Its activation also increases AMPA receptor endocytosis at excitatory postsynaptic sites and induces LTD (Chávez et al., 2010). The overall contribution of TRPV1 activation is considered pro-convulsive, as the TRPV1 agonist capsaicin was found to aggravate PTZ-induced seizures (Manna and Umathe, 2012). Furthermore, TRPV1 knockout mice exhibit decreased mortality due to PTZ-induced seizures (Jia et al., 2015). Therefore, AEA could be pro-convulsive through activation of TRPV1.

Involvement of the CB₁ Receptor

Early studies repeatedly demonstrated the suppressive effects of the CB₁ agonists, Δ⁹-THC and WIN55212-2, on seizures (Boggan et al., 1973; Wada et al., 1975; Wallace et al., 2001). However, a protective effect of CB₁ receptor signaling against acute seizures was clearly demonstrated later using a kainate-induced acute seizure model in CB₁ knockout mice (Marsicano et al., 2003). A subsequent study demonstrated that CB₁ receptors in hippocampal excitatory neurons, but not those in inhibitory neurons, were necessary to suppress kainate-induced seizures (Monory et al., 2006) despite the fact that the expression of CB₁ mRNA in glutamatergic neurons were lower than in GABAergic interneurons (Marsicano and Lutz, 1999). Furthermore, overexpression of CB₁ receptors in hilar mossy cells of the dentate gyrus and in pyramidal cells of the hippocampal CA1, CA2 and CA3 regions were shown to significantly reduce the severity of kainate-induced SE (Guggenhuber et al., 2010) and the expression of CB₁ receptors in excitatory neurons of the cerebral cortex, hippocampus, and amygdala of global CB₁ receptor knockout mice was sufficient to prevent the exacerbation of kainate-induced seizures in these mice (Ruehle et al., 2013). Taken together, these results indicate that CB₁ receptor signaling at excitatory synaptic terminals in the hippocampus has a suppressive effect on kainate-induced seizures (Table 1).

In addition to kainate-induced seizures, endocannabinoid signaling mediated by CB₁ receptors also effectively suppresses pilocarpine-induced seizures (Kow et al., 2014; Carletti et al., 2015). Recently, it was reported that the suppressive effect of WIN55212-2 on pilocarpine-induced seizures in rats was further enhanced by the co-administration of the 5-HT_2B/2C receptor agonist RO60-0175, which RO60-0175 application alone did not affect seizures (Colangeli et al., 2019). Furthermore, the 5-HT_2B receptor antagonist RS127445 significantly blocked the effects of WIN55212-2 and RO60-0175 co-administration. Therefore, CB₁ and 5-HT_2B receptor signaling can synergistically suppress pilocarpine-induced seizures.

Seizures evoked by electrical stimulation of the amygdala were more severe in excitatory neuron-specific CB₁ knockout mice and milder in inhibitory neuron-specific CB₁ knockout mice than in wild-type mice (von Rüden et al., 2015b). These results are consistent with the notion that the activation of CB₁ receptors at excitatory synapses suppress, whereas those at inhibitory synapses disinhibit and aggravate seizures (Table 1).

In marked contrast to seizures induced by kainate, pilocarpine, or electrical stimulation, the severity of PTZ-induced seizures is similar between CB₁ knockout mice and their wild-type littermates (Naydenov et al., 2014). Some reports have shown both preventive (Naderi et al., 2011; Vilela et al., 2013) and promoting effects (Vilela et al., 2013) of CB₁ receptor signaling on PTZ seizures. As of now, it is difficult to determine whether CB₁ receptor signaling has anti- or pro-convulsive effects on PTZ-induced seizures and this issue requires further detailed investigation (Table 1).

The results described above were obtained from seizures in the non-epileptic brain. In rats with spontaneous seizures after pilocarpine-induced SE, the number of spontaneous seizures was significantly decreased by the CB₁ receptor agonist WIN55212-2 and significantly increased by the CB₁ receptor antagonist SR141716A (Wallace et al., 2003). Consistent results were obtained by using the CB₁ receptor blocker AM251 for spontaneous seizures after kainate-induced SE (Sugaya et al., 2016) and by using positive allosteric modulator of CB₁ receptor for the spontaneous absence seizures in GAERS rats (Roebuck et al., 2021). These results indicate that CB₁ receptor signaling in the epileptic brain suppresses the occurrence of spontaneous seizures in the SE model and the absence seizure model (Figure 2 and Table 1).

WIN55212-2 injection decreased the number of spontaneous spike-and-wave discharges for 3 h in WAG/Rij rats. Interestingly, very long, sporadic trains of spike-and-wave discharges were observed in WAG/Rij rats from 3 h after WIN55212-2 injection (van Rijn et al., 2010), which is in sharp contrast to the decreased duration of spike-and-wave discharges after CB₁ activation in GAERS rats (Roebuck et al., 2021). These long spike-and-wave discharge trains were partially suppressed by co-administration of AM251 (van Rijn et al., 2010), suggesting the involvement of CB₁ receptor signaling in these abnormal paroxysmal discharges. Therefore, CB₁ receptor signaling potentially decreases the number of absence seizures but increases the duration of each seizure in WAG/Rij rats.