Wistar male rats (n = 766) weighing 250–300 g were used. They were kept under 20 °C–22 °C, humidity levels of 50%–60% and 12-h light/dark cycle. Food and water were provided ad libitum. Animals were habituated to manipulation by receiving a daily administration of saline solution (1 mL/kg, i.p.) during 7 days. Habituation was carried out to avoid the influence of stress in the results obtained because it mitigates changes on blood pressure, heart rate (Bonnichsen et al., 2005), drug side effects (Abbott et al., 2006), nociception (Smythe et al., 1994) and sensorimotor gating (Santos-Carrasco et al., 2025) induce during the experimental procedure. The present study adhered to the Mexican Official Standard (NOM-062-ZOO-1999) and received approval from the Ethics Committee of the Center for Research and Advanced Studies (CICUAL 0394-24).

Initially, a group of animals (n = 300) previously habituated to manipulation (see Section 2.1) was used to estimate the ED₃₀ and ED₅₀ of phenobarbital, diazepam, valproic acid, lamotrigine, and oxcarbazepine to prevent severe clonic-tonic seizures (wild jump, running and clonic-tonic seizures) induced by an acute administration of 3-MP. For this purpose, animals pretreated with one ASM (n = 10 per dose per ASM) were treated with 3-MP (37.5 mg/kg i.p.). The interval between the administration of the ASMs and 3-MP was 1 h, except for diazepam (2 h). Thereafter, the expression of severe generalized seizures was estimated during 30 min. To estimate the dose-response curve, each ASM had at least six doses evaluated (dose ranges for each drug are presented in Table 1). The data were analyzed using nonlinear regression to determine the ED₃₀ and ED₅₀ of each ASM (Motulsky and Christopoulos, 2003).

A second experiment focused on determining if repetitive administration of CBD combined with ASMs induces side effects on sensorimotor function that may confound anticonvulsant effects. For this experiment, rats previously habituated (see Section 2.1) received CBD (200 mg/kg p.o.), or coconut oil (CCO, 9.5 mL/kg p.o.). One hour later, they received an ASM at ED₃₀ or ED₅₀, or vehicle (saline solution 1 mL/kg i.p.). We used 3 rats per each treatment (a total of 66 rats). This procedure was repeated every 12 h during 10 trials. The sensorimotor function was evaluated under basal conditions (without treatment), and after the first, fifth, and 10th trials as well as 48 h after the last administration. The interval between the administration of the ASM and the evaluation of the sensorimotor function was 1 h except for diazepam (2 h).

The sensorimotor function was estimated using Neuroscore (Pierce et al., 1998), a battery of tests that evaluates different motor responses such as the ability to remain on an inclined surface at different angles (35°–75°) in both vertical and lateral positions (left (L) and right (R)), hind and forelimb contraflexion (L-R) during tail suspension, and resistance to lateral displacement (L-R). A score of 26–28 indicates normal function, whereas lower scores suggest sensorimotor impairment. The sensorimotor results were assessed by two-way ANOVA post hoc Tukey test.

This experiment was designed to determine if the administration of CBD in combination with ASMs at ED₃₀ or ED₅₀ could have an impact on the mortality rate and development of drug-resistant seizures as consequence of the repetitive administration of 3-MP. Remarkably, repetitive administration of 3-MP is known to induce over-expression of P-glycoprotein in the brain and drug-resistance to phenytoin and phenobarbital in rodents (Lazarowski et al., 2004; Enrique et al., 2017; Pérez-Pérez et al., 2021).

For the present study, we used the protocol previously reported by Pérez-Pérez et al. (2021) to induce drug-resistant seizures. Rats (n = 400) were habituated as previously described (see Section 2.1). One day after finishing the habituation, a group of animals was used for evaluating the mortality rate and the expression of phenobarbital-resistant seizures after receiving the subchronic coadministration of CBD plus phenobarbital. For this purpose, different experimental groups were designed to receive one of the following treatments:

a. CBD plus phenobarbital at ED₅₀. CBD (200 mg/kg p.o.) was delivered 1 h before phenobarbital. 1 h after phenobarbital administration, 3-MP was applied at an initial dose of 30 mg/kg, i.p. This procedure was repeated every 12 h during 10 trials. The 3-MP dose was gradually increased through the experimental procedure up to 37.5 mg/kg to maintain the induction of severe clonic-tonic seizures (Enrique et al., 2017). The mortality rate was evaluated during the repetitive administration of 3-MP. 12 h after the last 3-MP administration, an assay focused on identifying animals with phenobarbital-resistant seizures was carried out in 10 surviving rats. The assay consisted in the administration of phenobarbital and 60 min later, a proconvulsant dose of 3-MP (37.5 mg/kg). Animals showing clonic-tonic seizures during this assay were considered resistant to phenobarbital.

Similar treatments were designed with diazepam, valproic acid, lamotrigine and oxcarbazepine to identify rats with drug-resistant seizures to these ASMs.

The total number of animals used for each treatment depended on the mortality rate (see Table 3). The mortality rate during the repetitive administration of 3-MP and the prevalence of animals with drug-resistant seizures at the end of the experimental protocol were analyzed employing Fisher’s exact test (one-side) and using GraphPad Prism version 8 for Windows (GraphPad Software, Inc.). A p < 0.05 was considered significant.

The interactions between sodium channels and CBD, lamotrigine and oxcarbazepine were studied via docking simulations. Autodock-GPU and Autodock Vina were considered to identify which software was best suited for this particular system.

The docking regions were defined from recent structural studies based on cryo-EM structures of human NaV1.7 by Huang et al. (2023a) and Wu et al. (2023), who described two binding regions apart from the central cavity of the pore domain. To evaluate whether the binding modes in NaV1.7 of CBD and classic antiseizure medications based on blockade of sodium currents may be reasonably extrapolated to other channel subtypes, a structural comparison between the binding regions of NaV1.7 and the isoforms associated with seizure control (NaV1.2 and NaV1.6) was performed by sequence-structure alignment tools in UCSF Chimera, using Needleman-Wunsch algorithm with BLOSUM-62 matrix for the alignment. The alpha subunits were retrieved from PDB (PDB-IDs 8G1A for NAV1.7, 6J8E for NaV1.2, and 8FHD for NaV1.6). The analysis was restricted to a radius of 10 Å from the centers of the binding regions, which were defined by the positions of the CBD (I-site) and carbamazepine (BIG-site) in the experimental structures. The high levels of sequence identity in the α-subunit for all isoforms, and in the BIG and I-sites (>80%, Figure 1) supported the extrapolation of the observations reported for this receptor to epilepsy-related NaV isoforms.

The control group receiving vehicle before the acute administration of 3-MP showed a prevalence of 100% of severe clonic-tonic seizures. The dose-response curve obtained from the different doses evaluated from each ASM allowed the estimation of the ED₃₀ and ED₅₀ to prevent severe clonic-tonic seizures (Table 1).

Rats receiving the repetitive administration of coconut oil plus vehicle showed an average Neuroscore index of 27.06 ± 0.18, indicating normal sensorimotor function. Similarly, the repetitive administration of CBD alone or in combination with the different ASMs at ED₃₀ or ED₅₀ also yield a Neuroscore index in the range of 26–28. These results support that the repetitive coadministration of CBD with the different ASMs at ED₃₀ or ED₅₀ did not induce sensorimotor side effects (Table 2).

Experimental groups receiving coconut oil plus vehicle during the repetitive administration of 3-MP showed an elevated mortality rate. In contrast, coadministration of CBD and phenobarbital at ED₃₀ or ED₅₀ during the repetitive 3-MP administration resulted in 0% of mortality (p = 0.0163 vs coconut oil plus vehicle for both doses). Similarly, coadministration of CBD and diazepam at ED₅₀ avoided the mortality (0%, p = 0.0433 vs. coconut oil plus vehicle). On the other hand, a high mortality rate (41%) was obtained when animals received the subchronic administration of CBD plus oxcarbazepine at ED₅₀. Other treatments did not significantly modify the mortality rate during the repetitive administration of 3-MP (Table 3).

The drug-resistant assay carried out by the end of the experimental procedure revealed the high prevalence of drug-resistant seizures to the different ASMs in the experimental groups receiving repetitive coconut oil plus vehicle. Low prevalence of animals with phenobarbital-resistant seizures was found when applied CBD plus phenobarbital at ED₃₀ (p = 0.005 vs coconut oil plus phenobarbital at ED₃₀) or ED₅₀ (p = 0.009 vs. coconut oil plus phenobarbital at ED₅₀). Similarly, coadministration of CBD and diazepam at ED₅₀ reduced the prevalence of diazepam-resistant seizures (p = 0.0286 vs. coconut oil plus diazepam at ED₅₀). Contrariwise, high prevalence (100%) of resistant seizures to oxcarbazepine (p = 0.0163 vs. coconut oil plus vehicle) was observed when animals received the subchronic administration of CBD plus oxcarbazepine at ED₅₀. No significant changes were detected with other treatments (see Table 4).

Recent studies (Huang et al., 2023b; Huang et al., 2023a; Wu et al., 2023) provided structural information about the interaction of CBD and different anesthetic drugs and ASMs (including carbamazepine and lamotrigine) with the NaV1.7 isoform of human sodium channels. They identified two binding sites for CBD, one in a fenestration in the upper pore region (IV-I fenestration, F-site) and the other close to the Isoleucine-Phenylalanine-Methionine (IFM) motif near the intracellular linker between domains III and IV (I-site). We found that CBD induces important conformational changes in the I-site, mainly in the S6III region, which stabilizes the inactivated state of the channel via an allosteric effect (Figure 2). In the case of lamotrigine, a dual-pocket inhibition was proposed (Figure 3), with one binding site in the central cavity of the pore domain (C-site), and a second one beneath the intracellular gate (BIG site). This same binding region was previously identified for carbamazepine (Wu et al., 2023).

Binding of lamotrigine promotes a conformational change characterized by a α→π transition in the middle of the S6IV region and, to a lesser extent, movements of S6II and S6III. Figure 4 shows the superimposition of the I- and BIG sites of NaV1.7, in the presence of CBD or lamotrigine, respectively. The binding sites are contiguous and delimited by the S6IV domain, which changes its position to accommodate these ligands. Some amino acids are shared by the two binding regions. The results obtained indicate conformational changes induced by the presence of CBD in the I site that affect the formation of the BIG site and prevent the interaction with lamotrigine.

At present, there is no structural data available about the interaction of oxcarbazepine with sodium channels. On the other hand, the high molecular similarity between carbamazepine and oxcarbazepine suggests that both drugs are likely to interact in a similar manner with the sodium channels. Then, we performed docking simulations considering the BIG site, a region in which carbamazepine interacts (Wu et al., 2023). Re-docking validations (Figure 5) showed that Autodock Vina was more accurate to predict the experimental binding pose. Figure 6 shows the binding mode proposed for the active metabolite of oxcarbazepine, (S)-(+)-10-Hydroxy-10,11-dihydrocarbamazepine (S-MDH). Interestingly, docking simulations of lamotrigine and S-MDH into the conformation induced by CBD to NaV1.7 (PDB-ID 8G1A) showed that the binding region of these ASMs is modified upon CBD binding (Figure 5).